F I S C A L A F F A I R S D E P A R T M E N T

Former Yugoslav Republic of Macedonia Fiscal Transparency Evaluation

Amanda Sayegh, Suzanne Flynn, Alberto Soler, Phillip Stokoe, and Ali Uppal

October 2018

3

CONTENTS

GLOSSARY _______________________________________________________________________________________ 5

PREFACE _________________________________________________________________________________________ 6

EXECUTIVE SUMMARY __________________________________________________________________________ 7

I. FISCAL REPORTING __________________________________________________________________________ 12 1.1. Coverage of Fiscal Reports __________________________________________________________________ 13 1.2. Frequency and Timeliness of Fiscal Reporting _______________________________________________ 21 1.3. Quality of Fiscal Reports _____________________________________________________________________ 22 1.4. Integrity of Fiscal Reports ____________________________________________________________________ 24 1.5 Recommendations ___________________________________________________________________________ 26

II. FISCAL FORECASTING AND BUDGETING __________________________________________________ 29 2.1. Comprehensiveness of Budget Documentation _____________________________________________ 30 2.2. Orderliness __________________________________________________________________________________ 38 2.3. Policy Orientation ___________________________________________________________________________ 39 2.4. Credibility ___________________________________________________________________________________ 42 2.5. Recommendations ___________________________________________________________________________ 44

III. FISCAL RISKS _______________________________________________________________________________ 47 3.1. Disclosure and Analysis ______________________________________________________________________ 48 3.2. Fiscal Risk Management _____________________________________________________________________ 51 3.3. Fiscal Coordination __________________________________________________________________________ 59 3.4 Recommendations ___________________________________________________________________________ 62

FIGURES 1.1. Coverage of Public Sector Institutions in Fiscal Reports _____________________________________ 15 1.2. General Government Other Accounts Payable in Selected Countries, 2016 __________________ 17 1.3. Public Sector Balance Sheet Coverage in Fiscal Reports, 2016 _______________________________ 18 1.4. Public Sector Gross Liabilities in Selected Countries _________________________________________ 19 1.5. Public Sector Net Worth in Selected Countries ______________________________________________ 19 1.6. Contributions to Changes in General Government Debt, 2008–17 ___________________________ 24 2.1. Medium-Term Macroeconomic Forecast Errors, 2008–17 ____________________________________ 32 2.2. Average Medium-Term Fiscal Forecast Error, 2008–17 _______________________________________ 34 2.3. Budget Forecast Errors for Capital Expenditure and Revenue ________________________________ 35 2.4. Evolution of Fiscal Forecast Errors ___________________________________________________________ 35 2.5. General Government Investment in Emerging Europe and Central Asia, 2016 _______________ 38 2.6. Components of Deviation in Outcome from Approved Budget ______________________________ 43 2.7. Revisions to Medium-term Plans, 2008–17 __________________________________________________ 44 3.1. Volatility of Nominal GDP and Revenue Growth, 2000–17 ___________________________________ 48 3.2. Old Age Dependency Ratios _________________________________________________________________ 51

4

3.3. Use of the Contingency Reserves ____________________________________________________________ 52 3.4. Central Government Debt Maturity Profile __________________________________________________ 53 3.5. Government Guarantees in Europe, 2016 ____________________________________________________ 55 3.6. Financial Sector Liabilities – Select Countries ________________________________________________ 57 3.7. Natural Resource Rents, 2016 _______________________________________________________________ 58 3.8. Average Annual Damages from Natural Disasters ___________________________________________ 59 3.9. Size and Self-Reliance of Sub-National Governments, 2016 _________________________________ 60 3.10. Sub-National Government Financial Position, 2017 ________________________________________ 61 3.11. Liabilities of Public Corporations in Select Countries _______________________________________ 62 3.12. Stock of Guaranteed Public Debt by PC ____________________________________________________ 62 TABLES 0.1. Summary Assessment Against the Fiscal Transparency Code ________________________________ 10 0.2. Public Sector Financial Overview, 2016 ______________________________________________________ 11 1.1. List of Fiscal Reports _________________________________________________________________________ 13 1.2. Public Sector Institutions and Finances, 2016 ________________________________________________ 14 1.3. Example Reconciliation Table ________________________________________________________________ 26 1.4. Summary Evaluation: Fiscal Reporting _______________________________________________________ 28 2.1. Fiscal Forecasting and Budget Documents __________________________________________________ 29 2.2. Dates of Budget Submission and Approval __________________________________________________ 39 2.3. Fiscal Targets and Objectives Compared to Outturns ________________________________________ 40 2.4 Summary Evaluation: Fiscal Forecasting and Budgeting ______________________________________ 46 3.1. Reports Related to Fiscal Risks _______________________________________________________________ 47 3.2. Selected Specific Fiscal Risks _________________________________________________________________ 49 3.3. Indicators of Banking Sector Stability, 2017__________________________________________________ 57 3.4. Explicit Financial Sector Exposure, 2016 _____________________________________________________ 57 3.5 Summary Evaluation: Fiscal Risks _____________________________________________________________ 64

5

GLOSSARY CBA Cost-Benefit Analysis CMC Crisis Management Centre DIF Deposit Insurance Fund ELEM JSC Macedonian Power Plants ERP Economic Reform Program ESA European System of National Accounts FAD Fiscal Affairs Department FSR Fiscal Strategy Report GFSM Government Finance Statistics Manual IMF International Monetary Fund MBDP Macedonia Bank for Development Promotion MEPSO Electricity Transmission System Operator of Macedonia MoE Ministry of Economy MoF Ministry of Finance MPT Macedonia Radio Television MTBF Medium-term Fiscal Framework MZT Macedonia Railways Infrastructure NBRM National Bank of the Republic of Macedonia PESR Public Enterprise for State Roads PC Public Corporation PFM Public Financial Management PPP Public-Private Partnership SAO State Audit Office SSF Social Security Fund SSO State Statistical Office

LEVEL OF PRACTICE

RATING Not Met Basic Good Advanced

LEVEL OF IMPORTANCE

RATING High Medium Low

6

PREFACE An IMF technical assistance mission was in Skopje from April 30 to May 14, 2018 to conduct a Fiscal Transparency Evaluation based on the IMF’s Fiscal Transparency Code. The mission comprised Amanda Sayegh (head) and Alberto Soler of the IMF’s Fiscal Affairs Department, Phillip Stokoe of the IMF’s Statistics Department, Suzanne Flynn (FAD regional advisor) and Ali Uppal (FAD expert). Suzanne Flynn participated in the mission during the period April 30 to May 4, 2018. A preliminary visit by Suzanne Flynn also took place on March 30, 2018 to discuss the fiscal transparency framework and data requirements.

In the conduct of the evaluation, the mission met with the Minister of Finance, Mr. Dragan Tevdovski and senior officials of the Ministry of Finance, including from the Budget and Funds Department, Treasury Department, Macro-Economic Department, Debt and International Relations Department and the Taxation and Customs Policy Department.

The mission also met with the Chairperson and representatives of the Budget Council of the Assembly, President of the Finance and Budget Committee, Parliamentary Institute, and senior officials from the Public Procurement Bureau, State Statistics Office, State Audit Office, National Bank of the Republic of Macedonia, Central Registry, Ministry of Economy, Ministry of Labor and Social Protection, Ministry of Health, Ministry of Transport, the Health Insurance Fund, the City of Skopje Administration, the Public Enterprise for State Roads and the Public Transport Corporation of Skopje. In addition, the mission met with representatives from the EU Delegation to Macedonia and the World Bank.

The evaluation is based on information available at the time of the visit in May 2018. The findings and recommendations of the report represent the views and advice of the IMF mission team and do not necessarily reflect those of the authorities. Unless otherwise specified, the data presented in text, figures and tables in the report are estimates made by the IMF mission team and not official estimates of the government of Macedonia.

The mission would like to thank the Macedonian authorities and other officials for their excellent collaboration in the conduct of this evaluation and for the open exchanges of views on all matters discussed. Particular thanks go to Ms. Biljana Minoska (Deputy Head of the Treasury Department), Ms. Biljana Davceva-Mijoski and Mr. Gjorgji Nacevski (both from the IMF Local Office) for their substantive input and ongoing support to the mission. The mission would also like to thank Rohini Ray (FAD Research Assistant) for her support in compiling data and cross-country comparisons. Finally, the mission would like to express its appreciation to Mr. Gordan Tanaskov and Ms. Barbara Dragovik for their excellent interpretation and translation services.

7

EXECUTIVE SUMMARY Macedonia is making a concerted effort to enhance its fiscal transparency practices. The government has placed improving the quality and transparency of public institutions at the core of its 2018–21 Public Financial Management (PFM) Reform Program. Some progress has been made, particularly on increasing the availability of information on the use of public resources during the year, and further initiatives are underway to strengthen the medium-term focus and policy orientation of the budget as well as the legal framework for budget formulation and reporting. Further compliance with the European Union’s fiscal reporting requirements, as a candidate country for EU membership, will also strengthen transparency.

An assessment of fiscal transparency practices against the IMF’s Fiscal Transparency Code, finds that some elements of good practice are in place.1 Macedonia meets the standard of good or advanced practice on 13 of the 36 principles, and the basic standard on a further 12 principles (Table 0.1). Moreover, in 5 of the areas where Macedonia’s transparency practices do not currently meet basic practice, this could be addressed by publishing data that are already collected for internal management purposes. Areas of strengths include:

• Publication of extensive and timely in-year reports on budget execution for central and local budgetary units, on an economic and administrative basis, as well as for the three social security funds;

• Presentation of medium-term macroeconomic and fiscal projections in budget documentation, combined with a clear legal framework for budget formulation, and timely submission of the draft budget to the National Assembly;

• Publication of information on several risks to the public finances, including detailed analysis of the impact of alternative macroeconomic scenarios and, extensive disclosure of public debt and explicit guarantees, accompanied by sound frameworks for their management.

At the same time, this evaluation highlights several areas where Macedonia’s fiscal transparency practices could be improved:

• While fiscal reports and statistics cover the bulk of general government activity (90 percent), they exclude some public units (with net expenditures of around 3 percent of GDP), that should be classified as extrabudgetary units in accordance with international standards;

• Although comprehensive data is published on general government debt obligations, fiscal reports do not include information on other liabilities or financial assets and there is no reporting on revenues forgone from tax expenditures;

1 For earlier assessments of fiscal transparency in Macedonia, see IMF, Macedonia: Report on Observance of Standards and Codes—Fiscal Transparency, 2006, and International Budget Partnership, Open Budget Survey, 2017. The 2006 IMF report is not directly comparable to this one as it was based on an earlier version of the Code.

8

• The State Audit Office does not offer an opinion on whether the government’s annual budget report fairly represent public finances, with the annual audit largely compliance based and focused only on the execution of the central government’s core budget;

• Differences in key fiscal aggregates between fiscal and statistical reports are not explained, and no report explains how the change in the government’s debt can be reconciled with the budget deficit, which is important given that in some years, there have been material changes in the debt stock that have arisen from factors other than deficit financing needs;

• Although the budget includes most general government revenues and expenditures, it excludes the share of tax revenues levied by the government and allocated to the Public Enterprise for State Roads (0.7 percent of GDP) and taxes on banks to meet potential deposit insurance claims (0.2 percent of GDP);

• Despite efforts to enhance the policy orientation of the budget, the budget remains largely input-based, and credibility is hampered by limited reporting on performance against fiscal objectives and the absence of reconciliation of differences between successive fiscal plans;

• While the budget presents three-year costs for public investment projects, their total costs are not disclosed and there is no requirement to conduct or publish cost-benefit analyses for projects financed domestically or through bilateral loans;

• Information on the most important sources of fiscal risks are presented in different documents, but no report provides a comprehensive picture of the range of risks that public finances are exposed to; and

• There is limited central oversight of public-private partnerships and concessions and limited central monitoring of public corporation performance.

Based on these findings, the evaluation provides several recommendations to enhance fiscal transparency in Macedonia. Specifically, it recommends that the authorities:

• Expand the institutional coverage of fiscal and statistical reports, to include the activities of all institutional units that would be classified as part of the general government under international statistical standards;

• Enhance the quality of fiscal reporting, by: publishing data on government assets and financial liabilities; reconciling changes in the stock of debt and budget deficits; and explaining differences in fiscal aggregates between statistical and fiscal reports;

• Publish regular information on estimates of revenues forgone from tax expenditures;

• Expand the coverage of the audit of the annual budget report by requiring audit of the entire budget including own source revenue, loans and grants and expanding the number of financial audits of entities undertaken by the State Audit Office;

• Improve the comprehensiveness of the budget by including those tax revenues allocated to the Public Enterprise for State Roads and Deposit Insurance Fund and their associated

9

transfers, and presenting complete information on the own revenues and expenditures of all extrabudgetary funds in summary tables of consolidated fiscal aggregates;

• Strengthen budget credibility and transparency by presenting multi-annual expenditure plans by administrative unit; reconciling changes in fiscal aggregates between successive fiscal forecasts, and encouraging independant evaluation of fiscal forecasts;

• Enhance the policy orientation of the budget by regularly reporting on compliance with medium-term fiscal objectives and enhancing reporting against performance indicators for key policy areas;

• Publish the total costs of investment projects and develop and publish methodological guidelines for the appraisal of large projects along with their assessments;

• Improve fiscal risk disclosure, by publishing a summary fiscal risks report that details all material risks to public finances, their magnitudes, and strategies for their management; and publishing long-term public debt projections; and

• Strengthen monitoring and oversight of fiscal risks arising from public private partnerships and public corporations.

Implementing these recommendations will enable the Macedonian government to present a more complete and informative account of its fiscal position, prospects, and risks. It would also provide a more comprehensive overview of the general government and public sector fiscal position. Based on 2016 data, including and consolidating extrabudgetary units under the control of government would add 1.8 and 3.1 percent of GDP to consolidated general government revenue and expenditure respectively, and would have the effect of increasing the size of the reported deficit by 1.3 percent of GDP in that year. Further, presenting more complete balance sheet information would show that the public sector has extensive assets amounting to 106.8 percent of GDP (of which 65.2 percent of GDP is nonfinancial assets and 41.6 percent of GDP is financial assets), liabilities of 79.8 percent of GDP (of which 50.2 percent of GDP is debt securities and loans), and an overall net worth position of 27 percent of GDP.

The remainder of this report provides a more detailed evaluation of Macedonia’s fiscal transparency practices against the standards of the Code. It is organized as follows:

• Chapter I evaluates the coverage, timeliness, quality, and integrity of fiscal reporting;

• Chapter II evaluates the comprehensiveness, orderliness, policy orientation, and credibility of fiscal forecasting and budgeting; and

• Chapter III evaluates arrangements for disclosure and management of fiscal risks.

10

Table 0.1. Macedonia: Summary Assessment Against the Fiscal Transparency Code

I. Fiscal Reporting II. Fiscal Forecasting & Budgeting

III. Fiscal Risk Analysis & Management

Coverage of Institutions Budget Unity Macroeconomic Risks

Coverage of Stocks Macroeconomic Forecasts Specific Fiscal Risks

Coverage of Flows Medium-term Budget Framework Long-term Fiscal Sustainability

Coverage of Tax Expenditures Investment Projects Budgetary Contingencies

Frequency of In-Year Reporting Fiscal Legislation Asset and Liability Management

Timeliness of Annual Accounts Timeliness of Budget Documentation Guarantees

Classification Fiscal Policy Objectives Public-Private Partnerships

Internal Consistency Performance Information Financial Sector

Historical Revisions Public Participation Natural Resources

Statistical Integrity Independent Evaluation Environmental Risks

External Audit Supplementary Budget Subnational Governments

Comparability of Fiscal Data Forecast Reconciliation Public Corporations

11

Table 0.2. Macedonia: Public Sector Financial Overview, 2016 (Percent of GDP)

Source: Government Finance Statistics, Annual Financial Statements, Quarterly Public Sector Debt Statistics, NBRM and SSO, and IMF staff estimates. Notes: Central Government includes consolidated data for the budgetary central government, social security funds and extrabudgetary units. Financial accounts data from the NBRM used in the above analysis are still at the preliminary stages of development, and, in particular there is a high degree of uncertainty around estimates of accounts payable.

Central Government

Local Governments

Consolidation Gen. Govt.

Consolidated General

GovernmentNonfinancial Financial Central Bank

TransactionsRevenue 28.6 4.9 -3.1 30.5 5.3 0.1 0.2 -0.6 35.4Expenditure 32.7 4.9 -3.1 34.5 4.4 0.1 0.3 -0.6 38.6

Expense 28.4 3.9 -3.1 29.3 3.9 0.1 0.3 -0.6 32.9Investment in Non-financial assets 4.2 1.0 0.0 5.2 0.5 0.0 0.0 0.0 5.8

Gross operating balance 0.2 1.0 0.0 1.2 1.4 0.0 -0.1 0.0 2.5Net lending/borrowing -4.0 0.0 0.0 -4.0 0.9 0.0 -0.1 0.0 -3.3

Stocks

Assets 53.7 25.4 -0.4 78.7 25.4 2.5 29.8 -29.5 106.8Nonfinancial 19.6 22.6 0.0 42.2 22.8 0.0 0.1 0.0 65.2Financial 34.0 2.8 -0.4 36.5 2.6 2.5 29.7 -29.5 41.6Loans 0.5 0.0 -0.4 0.1 0.1 2.1 0.9 -0.9 2.3Equity 22.1 1.3 0.0 23.4 0.1 0.0 0.0 -22.3 1.1

Liabilities 49.8 2.0 -0.4 51.5 25.6 2.5 29.9 -29.5 79.8Liabilities other than equity 49.8 2.0 -0.4 51.5 6.5 2.1 26.9 -7.2 79.8

Debt securities and loans 42.3 0.5 -0.4 42.4 2.9 2.1 3.8 -1.1 50.2Accounts payable 7.6 1.5 0.0 9.0 3.7 0.0 0.1 0.0 12.8

Equity 0.0 0.0 0.0 0.0 19.0 0.4 3.0 -22.3 0.0Net worth 3.8 23.4 0.0 27.2 -0.2 0.0 0.0 0.0 27.0Net financial worth -15.8 0.8 0.0 -15.0 -23.0 0.0 -0.2 0.0 -38.2

General Government Public Corporations

Eliminations for Consolidation

Public Sector

12

I. FISCAL REPORTING 1. Fiscal reports should provide a comprehensive, timely, reliable, comparable, and accessible summary of the government’s financial performance, financial position, and cash flows. This chapter assesses the quality of Macedonia’s fiscal reporting practices against the standards set by the IMF’s Fiscal Transparency Code for the following dimensions:

• Coverage of public sector institutions, stocks, and flows;

• Frequency and timeliness of reporting;

• Quality, accessibility, and comparability of fiscal reports; and

• Reliability and integrity of reported fiscal data.

2. Macedonia publishes a range of fiscal information through a number of different documents. Macedonia’s main summary fiscal reports, presented in Table 1.1, comprise:

• Monthly budget execution reports, which present cash-based outturns for revenue, expenditure, and financing for the central government and the three Social Security Funds (SSFs)2 by detailed economic classification. In addition, monthly reports present cash-based budget execution for each budget user by economic classification;

• Monthly and quarterly economic reports, which include a discussion comparing outturns for the central government by economic classification with performance to the corresponding period of the previous year, and since July 2017 have expanded this to include a discussion on outturns for the largest budget users. Quarterly reports also include a discussion on general government debt trends;

• Quarterly budget execution reports, which present cash-based outturns for revenue, expenditure, and financing for the general government, but at a higher-level of aggregation than monthly execution reports;

• Quarterly debt statistics, which provide information on the stock of general government and guaranteed debt, including historical data by instrument;

• Semi-annual reports on budget execution, which present cash-based outturns for revenue, expenditure, and financing for the central government by budget user and the SSFs, and include commentary on spending and revenue performance over the first half of the year, as well as trends in general government and guaranteed debt;

• Annual report on the budget, which presents final outturns for revenue, expenditure, and financing for the central government by economic, functional, administrative and program classification, and for individual local governments by detailed economic classification; and

2 The Pension and Disability Fund, the Health Insurance Fund and the Employment Agency.

13

• Annual report on public debt management, which provides information on the stock of general government and guaranteed debt, including historical data by instrument.

Table 1.1. Macedonia: List of Fiscal Reports

REPORT Author COVERAGE ACCOUNTING PUBLICATION

Inst. Flows Stocks Basis Class. Freq. Lag IN-YEAR REPORTS

Monthly Budget Execution Tables

Treasury BCG, SSFs

R ,E, Fin

… Cash Nat. Monthly 25d

Monthly Budget Execution by Budget User

Treasury BCG, SSFs

R, E, Fin

… Cash Nat. Monthly 2m

Quarterly Budget Execution Tables

Treasury GG R, E, Fin

… Cash Nat. Monthly 1m

Short-Term Economic Trends

MD BCG R, E, Fin

… Cash Nat. Monthly 1m

Quarterly Economic Report

MD GG R, E, Fin

Debt Cash Nat. Quarterly 3m

Semi-annual report on Execution of the Budget

MD BCG, SSFs

R, E, Fin

Debt Cash Nat. Semi-annual

1m

Semi-annual Report on Local Governments

Treasury LG R, E …. Cash Nat. Semi-annual

1m

Quarterly Debt Statistics DIRD GG …. Debt,

Guarantees - Nat. Quarterly 30d

ANNUAL REPORTS

Annual Report on Budget BFD GG R ,E, Fin

… Cash Nat. Annual 6m

Government Financial Statistics (GFS)

Treasury GG R, E, Fin

Debt Cash GFSM2014 Annual 12m

Annual Report on Public Debt Management

DIRD GG R, E, Fin

Debt, Guarantees

- Nat. Annual 6m

Note: BCG: Budgetary Central Government; LG: Local Government; GG: General Government (which includes, BCG, SSFs, and LG); SSF: Social Security Funds; R: Revenue; E: Expenditure; Fin: Financing. MD: Macroeconomic Department, DIRD: Debt and International Relations Department; BFD: Budget and Funds Department.

1.1. Coverage of Fiscal Reports 1.1.1. Coverage of Institutions (Basic)

3. Macedonia’s public sector comprises 1,328 separate units of various legal forms. As shown in Table 1.2, these are distributed in the following subsectors:

• Central government, which is comprised of 398 budgetary central government units, including the central government administration and ministries and the three SSFs, and 131 extrabudgetary units, which include 113 public health institutions, 6 regulatory agencies3 and 12 public enterprises which should be classified inside the general government boundary. The largest of these units are:

3 These include, Agency for Electronic Communications, Macedonia Postal Agency, Agency for the Regulation of the Railway Sector, Agency for Audio and Audiovisual Media Services, Energy Regulatory Commission, Civil Aviation Agency.

14

• The Public Enterprise for State Roads (PESR), which had expenditures of 2.5 percent of GDP in 2016;

• Macedonia Railways Infrastructure (MZT), which had expenditures of 0.2 percent of GDP in 2016;

• Macedonia Radio Television (MPT), which had expenditures of 0.2 percent of GDP in 2016;

• The Deposit Insurance Fund (DIF), which had relatively little expenditures, but revenues of 0.2 percent of GDP and significant financial assets of 2.4 percent of GDP at end-2016.

• Local government, which is comprised of 668 units including 81 municipalities, and 587 other government bodies which are controlled by municipalities.

• Public nonfinancial corporations, which comprise 126 corporations, of which 21 are controlled by central government units, and 105 are controlled by local government units.

• Public financial corporations, which comprise the National Bank of the Republic of Macedonia (NBRM), the Macedonia Bank for Development Promotion (MBDP)4 and three financial regulatory agencies.5

Table 1.2. Macedonia: Public Sector Institutions and Finances, 2016 (Percent of GDP, unless otherwise stated)

Source: Authorities and IMF Staff Estimates. Note: ELEM refers to JSC Macedonian Power Plants and MEPSO refers to the Electricity Transmission System Operator of Macedonia.

4 The MBDP is classified by the authorities as a public financial corporation. While it is undoubtedly a public sector unit, questions remain on its classification as a general government unit or as a public financial corporation. Statistical guidance on the delineation of public units engaged in financial activities and general government units is not well defined in the IMF GFS Manual 2014 or in the European System of Accounts 2010, but the blanket government guarantee on liabilities of the MBDP enshrined in the establishing legislation may call into question classification of MBDP as a financial intermediary applying GFSM 2014 principles. The Irish government has established a similar entity called the Strategic Banking Corporation of Ireland, the Irish authorities, following advice from Eurostat classify this unit as a general government unit. (see http://ec.europa.eu/eurostat/documents/1015035/6761701/Advice-2015-IE-Classification-of-Strategic-Banking-Corporation-of-Ireland.pdf) 5 These include: The Agency for Supervision of Fully Funded Pension Insurance (MAPAS), Securities Exchange Commission and Insurance Supervision Agency.

Number of entities

Revenue Expenditure BalanceIntra-PS

expenditureNet expenditure

Net expenditure(Percent)

Public Sector 1,328 35.4 38.6 -3.3 0.0 38.6 100.0General government 1,197 30.5 34.5 -4.0 34.5 89.3

Central government 529 28.6 32.7 -4.0 3.1 29.6 76.6PESR 1.1 2.5 -1.4 2.5

Local governments 668 4.9 4.9 0.0 0.0 4.9 12.7City of Skopje 0.8 0.8 0.0 0.0 0.8

Nonfinancial public corporations 126 5.3 4.4 0.9 4.4 11.4ELEM 2.2 1.6 0.6 0.0 1.6 4.1MEPSO 0.9 0.6 0.3 0.4 0.2 0.6

Central Bank 1 0.2 0.3 -0.1 0.0 0.3 0.8Other financial public corporations 4 0.1 0.1 0.0 0.0 0.1 0.1

15

4. Macedonia’s public sector expenditures accounted for 38.6 percent of GDP in 2016. Table 1.2 summarizes the distribution of public resources across the different subsectors of the public sector and shows that:

• General government expenditure accounted for 34.5 percent of GDP, of which around 86 percent was spent by the central government and 14 percent was spent through local governments; and

• Public corporation expenditures accounted for 4.8 percent of GDP, almost all of which was spent by nonfinancial corporations.

5. While the government publishes a large volume of fiscal information, no report provides complete coverage of either general government or public sector activity. The monthly and annual fiscal reports for central government, and quarterly and annual reports for general government present fiscal data for the budgetary central government and local government sectors. While a substantial share of total general government activity is covered in these reports, they do not capture the activities of regulatory agencies (including the Agency for Electronic Communications and 5 smaller entities) and public corporations (PCs) that should be classified as extrabudgetary units on the basis of international statistical standards (GFSM 2014 and ESA 2010), nor do they include the activities of the public health institutions. Together the missing extrabudgetary units have net expenditures of around 3.2 percent of GDP (Figure 1.1).

Figure 1.1. Coverage of Public Sector Institutions in Fiscal Reports (Percent of total at each level)

Public sector expenditures Public sector liabilities

Source: Macedonian authorities and IMF staff estimates. Note: “Not Reported” refers to expenditures or liabilities of units not consolidated in summary fiscal reports.

6. More comprehensive coverage is provided by the quarterly debt statistics. This report includes liabilities in the form of debt securities and loans for the budgetary central government and local government and also the guaranteed debt of public enterprises.

16

7. Expanding the institutional coverage of fiscal reports to the entire public sector would have a substantial impact on many of the fiscal aggregates including the general government deficit. The reported general government deficit was 2.7 percent of GDP in 2015. Including and consolidating extrabudgetary units under the control of government would add around 1.8 percent of GDP to consolidated revenue and 3.1 percent of GDP to consolidated expenditure, which would have the effect of increasing the size of the reported deficit for the consolidated general government sector by 1.3 percent of GDP, to 4.0 percent of GDP. Expanding the coverage to the public sector by including PCs, would add 4.8 percent of GDP to revenue and 4.1 percent of GDP to expenditure, resulting in an overall public sector deficit of 3.3 percent of GDP.

1.1.2. Coverage of Stocks (Not Met)

8. Detailed information on public debt liabilities is published, but coverage of liabilities is incomplete, and no data is disclosed on nonfinancial or financial assets. The government debt reports present information on general government and government guaranteed debt in the form of debt securities and loans, which amounted to 48.4 percent of GDP at end-2016. The reports include the vast majority of public debt, but, do not account for certain liabilities, such as unguaranteed debt of public enterprises (around 0.2 percent of GDP) and NBRM bills (3.8 percent of GDP). Nor do they consolidate the debt holdings of the Deposit Insurance Fund (DIF). Accounting for these factors, would take public debt to 50.2 percent of GDP at end-2016. In addition, no fiscal or statistical reports currently provide information on government assets. Inclusion of government “currency and deposit holdings,” information on which is readily available, in the debt statistics and fiscal reports would see the assessment against this principle improve from not met to basic level of coverage.

9. Efforts are underway to compile broader balance sheet information and accounts. The NBRM has begun work on general government balance sheets as part of its wider work to compile financial accounts for the entire economy and expects to publish the results of this work once complete, although no precise timeframe has been set. In addition, the State Statistical Office (SSO) has done some preliminary work to estimate stocks of some general government nonfinancial assets but this has also not yet been published.

10. Although the coverage of public debt is broad, there are gaps in reporting of non-debt liabilities and assets (Figure 1.1). The main gaps reflect the following:

• Other accounts payable are not published. As Macedonia has cash-based accounting, the government is not yet able to easily compile information on payables. The Ministry of Finance (MoF) does collect information on committed expenditures each quarter and in May 2017 conducted a survey which found unconsolidated arrears across the entire public sector

17

of around 3.6 percent of GDP.6 But, it is not clear to what extent these reflect intra-public sector payments or overdue amounts. Preliminary analysis suggests that overall accounts payable (that is, unpaid invoices, irrespective of whether they are overdue), for the general government were in the order of 9 percent of GDP at end-2016, although there is a high degree of uncertainty around these estimates. While these estimates are slightly above the average across countries that report these data in GFS, they are not out of line with neighboring countries in South East Europe (Figure 1.2).

• NBRM liabilities are excluded from public sector debt. Although this includes technical liabilities, including currency in circulation (5.7 percent of GDP), SDR liabilities (2.7 percent of GDP), and banking system reserves (8.4 percent of GDP) the NBRM is also one of a minority of central banks that issues its own bonds. In most other countries, including all of Macedonia’s neighbors, central banks rely on government securities for monetary policy. The outstanding stock of NBRM securities were around 3.8 percent of GDP at end-2016.

• Nonfinancial and financial assets are not published in fiscal reports, although government deposits are disclosed in monetary data compiled by the NBRM. Based on available information, the general government is estimated to have nonfinancial assets of at least 42.2 percent of GDP, and equity in its PCs of around 23.4 percent of GDP. Macedonia’s equity assets include a 34.8 percent ownership stake in Macedonia Telecom, which was worth almost 1 percent of GDP in 2016 as well as a 50 percent ownership stake in the gas transmission system operator, GA-MA, estimated to be worth around 0.2 percent of GDP.

Figure 1.2. General Government Other Accounts Payable in Selected Countries, 2016 (Percent of GDP)

Source: IMF GFS Database and IMF Staff Estimates for Macedonia.

6 The survey suggests accounts payable of about 1 percent of GDP for central government budgetary units and a further 0.5 percent of GDP for public health institutions, 0.8 percent of GDP for local government units, and around 1.4 percent of GDP for central and local government enterprises. The Health Insurance Fund has estimated that public health institutions have arrears of 0.8 percent of GDP while the Fund itself has estimated accounts payable of 0.2 percent of GDP arising from lengthy payment terms for some health fund expenditures.

0

2

4

6

8

10

12

14

16

18

20

Cana

daPo

land BiH

Hung

ary

Alba

nia

Fran

ceM

alta

Gree

ceAu

stra

liaM

aced

onia

Switz

erla

ndN

ew Z

eala

ndPo

rtug

alDe

nmar

kCz

ech

Rep.

Rom

ania

Bulg

aria

Aust

riaFi

nlan

dU

.S.

Swed

enSl

oven

iaSl

ovak

Rep

.N

orw

ayIc

elan

dLu

xem

bour

gCy

prus

Italy

Spai

nN

ethe

rland

sLi

thua

nia

Latv

ia UK

Rep.

Kor

eaEs

toni

aKo

sovo

Turk

eyBe

lgiu

mIre

land

Germ

any

18

11. Addressing these gaps and expanding the balance sheet to the public sector would provide a more comprehensive view of public finances. As shown in Table 0.2 and Figure 1.3, consolidated public sector asset holdings and liabilities are estimated to have been around 106.8 percent of GDP and 79.8 percent of GDP, respectively, at end-2016. Public sector net worth and net financial worth are estimated to have been 27 percent of GDP and -38.2 percent of GDP. The main components include:

• Nonfinancial assets of 65.2 percent of GDP, which primarily comprises infrastructure assets (roads, railways, and utilities) and forestry assets.

• Financial assets of 41.6 percent of GDP, the bulk of which comprise assets of the NBRM.

• Liabilities of 79.8 percent of GDP, which primarily comprise government debt securities and loans, and NBRM liabilities including currency in circulation and banking system reserves.

Figure 1.3. Public Sector Balance Sheet Coverage in Fiscal Reports, 2016 (percent of GDP)

Source: IMF staff estimates, Macedonian authorities.

12. The extent and value of budgetary central government and local government nonfinancial assets is not known. Financial accounts for public enterprises, including full balance sheets, are produced which enable values to be estimated for the national road network, railway infrastructure, utilities, and cultivated biological assets. However, the value of government buildings (including schools, hospitals, government offices) and government owned land, for both central government agencies and ministries, and municipalities are currently not reported. The State Statistical Office (SSO) is doing some work to produce statistical estimates of these capital stocks, but work has been limited to public roads, rail and research and development assets so far. Consequently, values of nonfinancial assets produced for this report are likely to be significantly underestimated, resulting in lower net worth than the true figure for Macedonia.

-100 -50 0 50 100

Net Worth

General Government

Nonfinancial Corporations

Financial Corporations

Central Bank

Public Sector

Reported

Unreported

Liabilities Assets

PSF

19

13. Macedonia’s net worth sits in the upper-range of other countries and gross liabilities are relatively low (Figure 1.4 and 1.5). This reflects the fact that Macedonia does not have any state-owned commercial banks and does not have any employment related pension schemes for government employees, instead government employees are covered by the existing social security and mandatory defined contribution pension system. For Macedonia, the rising costs of pensions as a result of its aging population are reflected as long-term fiscal pressures rather than balance sheet liabilities (see principle 3.1.3).

Figure 1.4. Public Sector Gross Liabilities in Selected Countries (Percent of GDP)

Source: IMF staff estimates, Fiscal Transparency Evaluations.

Figure 1.5. Public Sector Net Worth in Selected Countries (percent of GDP)

Source: IMF staff estimates, Fiscal Transparency Evaluations.

0

50

100

150

200

250

300

350

400

450

Port

ugal

Irela

nd UK

Finl

and

Aust

ria

Braz

il

Tuni

sia

Rom

ania

Alba

nia

Peru

Phili

ppin

es

Russ

ia

Turk

ey

Keny

a

Uga

nda

Geo

rgia

Mac

edon

ia

Moz

ambi

que

Gua

tem

ala

General Government

Public Corporations

-150

-100

-50

0

50

100

150

200

250

300

Port

ugal UK

Tuni

sia

Irela

nd

Aust

ria

Moz

ambi

que

Alba

nia

Gua

tem

ala

Keny

a

Braz

il

Rom

ania

Phili

ppin

es

Geo

rgia

Mac

edon

ia

Peru

Finl

and

Uga

nda

Turk

ey

Russ

ia

20

1.1.3. Coverage of Flows (Basic)

14. Fiscal reports are all compiled on a cash basis, with no accrual adjustments or accrual transactions. Fiscal reports cover cash revenues, expenditures, and financing. However, cash-based data do not capture accrued expenditures and payment arrears and also omit some important non-cash flows. In addition, some cash flows have been incorrectly recorded including central government on lending to municipalities and public enterprises, which is recorded as expenditure with an offsetting adjustment to revenue (see 1.4.3 for additional details).

15. Expenditures associated with the issuance of compensatory “structural bonds,” have not been appropriately accounted for in fiscal reports. These bonds are a collection of instruments that have been used by the government to compensate households for deposits lost during the breakup of Yugoslavia or as compensation for property seized without compensation as well as for the rehabilitation and privatization of some banks in the process of transition. These bonds have been correctly treated as liabilities in the stock of debt figures published by the MoF, however, they have not been correctly reflected in the deficit numbers, which should have recorded transactions as capital transfers above the line when these were issued. Issuance of structural bonds have averaged around 0.1 percent of GDP annually in the period from 2011–2015, but the impact was much higher in the past. Macedonia issued EUR 546.5 million of structural bonds in 2000, worth around MKD 33 billion at today’s exchange rate, or around 13 percent of GDP for that year.

16. Contributions to 2nd pillar pensions are included as revenue and expenditure in the authorities’ fiscal reports.7 These are amounts collected by the Public Revenue Office and which are ultimately transferred to private pension firms that administer the individual pension accounts in the 2nd pillar, which should be treated as financing transactions below the line. These are recorded correctly in the GFS data reported to the IMF GFS Yearbook database and in 2016 were approximately 0.8 percent of GDP. Whilst the current treatment in the authorities’ fiscal reports overstates both revenues and expenditures, neither treatment has implications for the budget deficit.

1.1.4 Coverage of Tax Expenditures (Not Met)

17. The government does not publish estimates of the revenue loss from tax expenditures. There is no legislative requirement to report revenues foregone from tax expenditures. However, when proposing legislative or regularly amendments, budget users are required to provide estimates of their fiscal implications to the Government. Estimates of individual tax expenditures may also be prepared, on an ex ante and ex post basis, setting out the fiscal implications of different measures, but these are generally not published. Activities

7 The 2nd pillar pension is a term for funded pension schemes. In Macedonia, these are defined contribution individual accounts held with private pension funds.

21

envisaged under the PFM Reform Program to strengthen forecasting analysis and capacity will also aid the development of more comprehensive reporting on tax expenditures.

18. The government provides a number of tax exemptions, but there is uncertainty around their size. The main tax expenditures are associated with 15 designated Technological-Industrial Development zones, where exemptions are provided to investors for corporate income tax and personal income tax for salaries of their employees for their first ten years as well as from payment on value-added tax and customs duties for goods, raw materials, equipment and machines. The NBRM estimated that the fiscal costs of the personal and corporate income tax exemptions amounted to MKD 1.1 billion (or 0.2 percent of GDP) in 2016. Additional tax expenditures are associated with tax preferences for voluntary pension savings, an excise tax exemption for wine, reduced rate property tax on the principal dwelling, reduced rate value-added tax for certain products, and zero or reduced-rate customs tax provided for under free trade agreements. There are currently no formal estimates of the size of these tax expenditures. Publishing an annual estimate on the revenue foregone from tax expenditures, along with details of their policy rationale and beneficiary groups would enhance transparency.

1.2. Frequency and Timeliness of Fiscal Reporting 1.2.1. Frequency of In-Year Fiscal Reporting (Advanced)

19. Monthly and quarterly budget execution reports provide a timely picture of unfolding fiscal developments. The monthly cash-based budget execution reports cover central government revenues, spending and financing (including for the three SSFs) and are published within 25 days of the end of month. These reports provide a breakdown of spending by economic classification and revenue source at the aggregate central government level. In addition, since July 2017, the government has published monthly execution reports of spending and revenue for each administrative unit, with a breakdown by economic category, although these are published with a lag of two months. Reports showing general government execution of revenues and expenditures have also been published on a quarterly basis, since July 2017, within two months of the end of the quarter. However, these reports are less detailed than the central government reports and no data for the separate subsectors are shown. The government plans to further enhance transparency by publishing all payments executed through the treasury on a daily basis.

20. Quarterly and annual public debt reports provide detailed information on government and wider public sector debt. The quarterly reports cover central and local government debt, as well as publicly guaranteed debt, and are published within one month of the end of the quarter. The reports capture almost all public sector debt in the form of debt securities and loans, other than debt securities issued by the NBRM and provide a detailed breakdown by instrument.

21. While extensive data is published on budget execution, it is presented across several different reports which can reduce accessibility. The monthly and quarterly execution

22

reports for the central and general government only provide financial information on flows. Explanations for spending trends, including comparison with previous periods, are provided in separate monthly reports on short-term economic trends, and economic quarterly reports, and from July 2017, have included analysis of fiscal trends for the major budget users. A semi-annual report on execution of the budget provides more detailed analysis and commentary, including outturns for the year to date and comparison with previous periods and summary of public debt. Semi-annual tables of the revenues and expenditures by municipality are also published, in excel format to allow analysis. Reducing fragmentation across all these different reports could enhance accessibility, provided it could be done without reducing the timeliness of core information on budget execution.

1.2.2 Timeliness of Annual Financial Statements (Advanced)

22. Cash-based annual budget execution reports are published before or around the end of June the following year, together with the audit report. The annual report includes spending by central budget user, the SSFs, and local governments by economic classification and source of funds. No information on stocks is presented. The report shows functional and sub-function classification of spending for the central government and for programs under each administrative unit. Outcomes for individual municipalities are also included in the annual report, with their spending broken down by economic classification and source of funds. Interim (unaudited) annual outturn reports are published on the MoF website within two months of the year end.

23. While modified cash-based annual financial statements are required for all budget users, these are not consolidated or consistently published. The law on accounting for budget users does not require a full balance sheet, only the presentation of financial assets and liabilities. This information does not form part of the annual budget execution report and is not consolidated. The balance sheet information provided is audited by the State Audit Office (SAO) only when the entity is selected for annual audit. Plans to move to a modified accrual accounting basis are outlined in the government’s 2018–21 PFM Reform Program, but there are currently no plans to move to International Public Sector Accounting Standards. Should such accounting reforms be implemented in the future, the scope of the audit would also need to be expanded to include accrual based financial statements, and the time required to complete the audit would likely need to be extended beyond the current six months.

1.3. Quality of Fiscal Reports 1.3.1. Classification (Advanced)

24. Fiscal reports present spending by administrative, economic, and functional classification, and provide some program information. The authorities compile and report annual GFS to the IMF Statistics Department in the GFSM 2014 framework. Revenue, expenditure, and financing data in the authorities’ budget and national reports are disseminated in a national presentation that broadly follows the GFSM 86 presentation. The Annual Report also provides a

23

full functional breakdown of expenditures that is fully aligned with United Nations’ Classification of Functions of Government. In addition, the Annual Report includes expenditures by program for each administrative unit, although there is scope for this presentation to be enhanced (see principal 2.3.2).

1.3.2. Internal Consistency (Basic)

25. Macedonia only publishes one of the three internal consistency checks on fiscal data required under the Code. The statistical reports provide a full reconciliation between the above-the-line fiscal deficit and the below-the-line financing, detailing the movement in cash, acquisitions of other financial assets and incurrence of financial liabilities. These are provided in gross terms, and net out so that net financing needs match the fiscal balance. There is no reporting on the holders of government debt that can be compared with debt issued.

26. There is no reconciliation between the increase in debt and debt issued in any given year. The government’s Economic Reform Program (ERP), which is an annual document prepared and published by EU candidate countries, includes an annex table which reports the change in the debt stock broken down by primary balance, interest, and stock-flow adjustment—that is, the increase in debt which is due to reasons other than budget deficits. But, it does not provide a reconciliation or explanation of the factors behind this adjustment. Doing so would enhance analysis, given the stock-flow adjustment has averaged (in absolute terms) 0.8 percent of GDP, annually, over the past decade (Figure 1.6).8 Furthermore, combining information on financing and the stock of debt in the fiscal reports would enhance transparency by facilitating better understanding of the impact of financing on government liabilities.

8 Between 2008 and 2017 the debt stock increased by 25.4 percent of 2017 GDP. Most of this (24.3 percent of GDP) can be explained by financing needs to fund budget deficits implying a stock-flow adjustment of around 1.1 percent of 2017 GDP over this period. The bulk of this can be explained by increases in cash and deposits and offsets from the proceeds from privatizations. Valuation effects due to exchange rate movements were almost insignificant as almost all general government debt is issued in domestic currency or in Euros (to which the Macedonian Denar is pegged). This leaves an unreconciled change in the debt stock of just 0.4 percent of 2017 GDP, although in some years (notably, 2008 and 2013), which could be reduced further utilizing existing information on on-lending and debt restructuring.

24

Figure 1.6. Contributions to Changes in General Government Debt, 2008–17 (Percent of GDP)

Source: IMF Staff Estimates.

1.3.3 Historical Revisions (Not Met)

27. Fiscal statistics are not revised. Annual submissions of GFS data to the IMF only commenced in 2017 and historical fiscal data published by the authorities is not revised. The recommendations of this fiscal transparency evaluation would result in meaningful revisions to the GFS data, which, if adopted by the government, should be disclosed. The inclusion of PESR, MZT, MPT, DIF and other units classified inside general government should be implemented across the time series, at least back to 2013 (when PESR was created) and the impact of these reclassifications should be disclosed along with detailed data supporting the decision.

1.4. Integrity of Fiscal Reports 1.4.1. Statistical Integrity (Basic)

28. Fiscal statistics are compiled by the MoF on a GFSM 2014 basis and reported to the IMF for inclusion in the annual GFS database. These reports are compiled by staff within the Treasury, and there is currently no standalone fiscal statistics unit within the MoF. Within the European Union in particular, good practice is to ensure a more transparent delineation between the producers of source data and statistical compilers and to provide some operational independence for the compiler of fiscal statistics.

29. The SSO are developing the reporting of fiscal statistics to Eurostat on an ESA 2010 basis, but this project remains ongoing and no data have yet been published. The SSO takes fiscal data from MoF and supplements this with additional data for extrabudgetary units and makes a number of accrual and other adjustments to development fiscal data in line with EU requirements. In addition, the SSO is an independent institution, under the Macedonian Law of Statistics, such that as the SSO’s fiscal statistics come on stream and begin to be published, then these statistics could be considered to meet the advanced standard for statistical integrity.

-4

-2

0

2

4

6

8

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

Primary deficitInterestPrivatizationCashResidualDebt change

25

1.4.2 External Audit (Basic)

30. The independent SAO publishes an opinion on the central government’s core budget execution report. The 2010 Law on State Audit provides for the independence of the audit office, including parliamentary appointment of the state auditor and deputy and approval for its budget. The Law on Budget (Article 52) requires the SAO to submit its annual report to the Parliament within six months of the end of the year. The law specifies that the report be based only on the execution of the central government’s core budget, which excludes spending financed from own revenues, grants or loans. This annual audit is compliance focused and there is no opinion as to whether the final accounts present a true and fair view. The audit is also limited to revenue and expenditures and does not cover assets and liabilities. The most recent audit reports on the core budget execution have been qualified, due in part to lack of compliance with the budget calendar specified in the law on Budget.

31. The SAO has authority to audit all public spending and can conduct regularity, compliance, and performance audits of public sector entities, but coverage is limited. Audits are planned annually based on risk, but due to capacity constraints, only a sample of entities are audited each year. In 2016 for example, 118 entities (out of a total of more than 1,000 public sector institutions) were audited by the SAO, and only 19 percent of these audits expressed unqualified opinions. The Law on Companies requires audits of the financial statements of all medium and large, listed and limited liability companies. In addition, they are required to register their approved financial statements within two months and their audited financial statements within six months, at the Central Registry, and these are available to the public for a fee.

1.4.3 Comparability of Fiscal Data (Basic)

32. Execution reports are produced on the same basis as the budget, but there are differences in fiscal data across fiscal and statistical reports that are not publicly reconciled. The annual, quarterly, and monthly budget execution reports present outturns on the same basis as the budget and provides clear information on how the government has performed in executing the budget. However, different recording of a number of flows lead to differences between the budget execution reports, and the GFS data reported to the IMF even though the underlying data is identical.

33. The differences in revenue and expenditure, though not the deficit, are large. In 2016, general government revenue and expenditures in GFS were approximately 1.6 percent of GDP lower than that presented in the annual report. These differences are driven by three main factors:

• Treatment of 2nd pillar pension contributions: Budget execution reports record contributions to 2nd pillar pensions as government revenue and expenditure that flows through the accounts of the Pension and Disability Fund. In the GFS these flows are excluded

26

from revenue and expenditure and recorded below-the-line as financing flows. In 2016, these flows were approximately 0.8 percent of GDP.

• Treatment of On-Lending: Budget execution reports record on-lending to municipalities and PCs as capital expenditures above the line. To ensure that these transactions do not affect the deficit, the authorities include an adjustment to non-tax revenues of an equivalent amount. In GFS data, on-lending is recorded as a financial transaction below-the-line.

• Treatment of repayments of historic loans: Budget execution reports record these inflows as non-tax revenues, while in GFS data, these inflows are recorded as financing flows below-the-line. This change in treatment results in the small difference (less than 0.1 percent of GDP) in the deficit between GFS and budget execution reports.

34. Ahead of reporting GFS data to the IMF, the Treasury produced a reconciliation of the main differences between Budget execution reports and GFS for MoF management. This document, if published with additional details like those shown in Table 1.3 to fully explain the reasons for the differences, would enhance transparency.

Table 1.3. Example Reconciliation Table

2016

Annual Report GFSM 2014 Difference Revenues 181,745 171,873 -9,872

Minus 2nd Pillar Pension Contributions …. Minus on lending adjustment …. Minus repayments of historic loans ….

Expenditures 197,657 188,227 -9,430 Minus 2nd pillar pension contributions …. Minus on lending to public enterprises ….

Fiscal Balance -15,912 -16,354 -442 Source: Authorities, IMF Staff.

1.5 Recommendations 35. Macedonia’s fiscal reports contain some good elements. The assessment against the Code, summarized in Table 1.4, shows that the monthly budgetary central government reports are timely and provide detailed information on revenues, expenditures, and financing. The quarterly reports for general government budget execution and debt also provide substantial information on revenues, expenditures, and financing for the whole of budgetary general government, as well as almost all liabilities, in the form of debt securities and loans, for the Macedonian public sector. The annual report and the semi-annual execution reports also provide a good breakdown of revenues and expenditures by administrative, economic, and functional classifications.

36. However, there is considerable scope to enhance the coverage, quality, and integrity of fiscal and statistical reports. Fiscal reports provide an incomplete picture of

27

general government as not all institutions are covered in fiscal and statistical reports. Reports do not capture accrued expenditures that have resulted in mounting payment arrears in some periods, and some budgetary central government transactions are not correctly recorded. Furthermore, no information on government assets is disseminated, and there are gaps in the coverage of liabilities. Revenues forgone from tax expenditures are also not reported. And, differences in revenues and expenditures across different reports and between budget financing needs and changes in the stock of debt are not explained.

37. Based on the above assessment, the evaluation highlights the following priorities for improving the transparency of fiscal reporting:

• Recommendation 1.1: Expand the institutional coverage of fiscal and statistical reports. Adopt sectoral classifications for fiscal and statistical reports that align with international standards:

– Adopt in the Budget Law, internationally accepted criteria for determining the classification of institutional units for fiscal reporting purposes and publish the list of general government and public sector units; and

– Record the revenues, expenditures, and financing of extrabudgetary units not currently included in the GFS reports and report liabilities of these entities as general government debt in the quarterly and annual fiscal reports.

• Recommendation 1.2: Enhance the quality of fiscal reporting.

a. Publish data on assets and all liabilities in fiscal reports; and

b. Publish reconciliations of changes in the stock of debt and net financing in fiscal reports and include a reconciliation table in the central government annual financial statements to reconcile differences between it and other fiscal reports.

• Recommendation 1.3: Regularly report on revenues foregone from tax expenditures. Publish an annual estimate on the revenue foregone from tax expenditures, including details of the nature of these tax expenditures, their objectives, and the groups or sectors benefiting from them.

• Recommendation 1.4: Expand the coverage of the annual audit. Amend the Law on Budget (Article 52) to require audit of the entire budget including own source revenue, loans, and grants and expand the number of financial audits of entities undertaken and plan audits over the medium term to enhance coverage.

28

Table 1.4. Macedonia: Summary Evaluation: Fiscal Reporting Principle Assessment Importance Recs

1.1.1 Coverage of Institutions

Basic: Fiscal reports cover most general government units, but exclude some extrabudgetary units that should be classified within that sector under international standards.

High: Revenues and expenditures related to activities of extrabudgetary units are not included in general government fiscal reports, resulting in an understatement of expenditures and the budget deficit of 3 and 1.3 percent of GDP respectively in 2016.

1.1

1.1.2 Coverage of Stocks

Not Met: The government publishes detailed information on public sector debt liabilities, but no information is provided on other accounts payable, financial or nonfinancial assets.

Medium: Unreported public sector liabilities of around 31 percent of GDP and assets of 107 percent of GDP. 1.2

1.1.3 Coverage of Flows

Basic: Fiscal reports cover cash revenues, expenditures, and financing.

Medium: Cash recording potentially understates expenditures and prevents a fuller understanding of accounts receivable and payables.

1.1

1.1.4 Coverage of

Tax Expenditures

Not Met: Revenues foregone from tax expenditures are not reported.

Medium: Several tax exemptions exist, with uncertainty about their size. 1.3

1.2.1 Frequency of

In-Year Reporting

Advanced: Monthly central government execution reports are published within one month of the end of the period.

Medium: Expenditure outturns have deviated by around 2 percent of GDP on average from the original budget.

1.2.2

Timeliness of Annual

Financial Statements

Advanced: Cash based annual budget execution reports are published before or around the end of June the following year, together with the audit report.

Medium: Annual execution reports do not include information on stocks. Central government arrears were estimated at 1 percent of GDP in 2017.

1.3.1 Classification

Advanced: Fiscal reports include administrative, functional and economic classifications broadly consistent with international standards, as well as some program information.

Low: Extensive information is provided on the use of public funds, but program information is not clearly linked to policy objectives.

1.3.2 Internal Consistency

Basic: Fiscal reports include only one reconciliation of the budget balance and financing.

Medium: Annual differences between changes in the debt stock and budget financing has averaged around 0.8 percent of GDP in absolute terms.

1.2

1.3.3 Historical Revisions

Not met: Fiscal statistics are not revised.

Low: The priority is to prepare robust fiscal statistics.

1.4.1 Statistical Integrity

Basic: Fiscal statistics are disseminated broadly in accordance with international standards.

Low: Although there are no independent checks on fiscal statistics, they are still being improved.

1.4.2 External Audit

Basic: An independent supreme audit institution publishes an opinion report on implementation of the budget, based on a limited audit sample of entities.

High: Only about 10 percent of public sector institutions were subject to audit in 2016. Of these audits, only 19 percent received an unqualified opinion.

1.4

1.4.3 Comparability of Fiscal Data

Basic: Budget execution reports are prepared on the same basis as the annual budget. Differences between fiscal reports and GFS are not explained.

Medium: Revenues and expenditures in fiscal and statistical reports differ by 5 percent, but differences in the deficit are low.

1.2

29

II. FISCAL FORECASTING AND BUDGETING 38. Fiscal forecasts and budgets should provide a clear statement of the government’s budgetary objectives and policy intentions, and comprehensive, timely, and credible projections of the evolution of the public finances. This chapter assesses the quality of Macedonia’s fiscal forecasting and budgeting practices against the standards set by the four dimensions of the fiscal transparency code:

• The comprehensiveness of the budget and associated documentation;

• The orderliness and timeliness of the budget process;

• The policy orientation of budget documentation; and

• The credibility of the fiscal forecasts and budget proposals.

39. A number of reforms have been initiated or implemented over the past two years, to strengthen the transparency of fiscal forecasting and budgeting. These include:

• Publication of a Citizen’s Guide to the 2018 Budget, summarizing the main economic and fiscal forecasts and new policy initiatives, combined with a concerted effort to engage key stakeholders as part of budget policy deliberations;

• Ongoing reform of the Budget Law aimed at improving the quality and transparency of the budget and of medium-term budgeting procedures; and

• Proposed amendments to the Law of Public Procurement to enhance the effectiveness and transparency of competitive tender processes.

Table 2.1. Macedonia: Fiscal Forecasting and Budget Documents

Document Agency Coverage Accounting Publication

Date Institutions Flows Stocks Basis Class

Annual Budget MoF BCG and

SSFs R, E, Fin

- Cash Nat Dec

Fiscal Strategy Report

MoF GG R, E, Fin

Debt Cash Nat Required by

May Economic

Reform Program Govt. GG

R, E, Fin

Debt Cash Nat Jan

Annual Report of the Budget

MoF GG R, E, Fin

- Cash Nat May

Annual Report on Public Debt Management

MoF GG - Debt and

guarantees - Nat. June

Note: BCG: Central Government; SSFs: Social Security Funds; GG: General Government (includes BCG, SSFs, LGs); R: Revenue, E: Expenditure; Nat: National.

30

2.1. Comprehensiveness of Budget Documentation 2.1.1. Budget Unity (Not Met)

40. The budget includes revenues and expenditures of most central government budgetary units on a gross basis, but does not include all tax revenues. The annual budget includes revenues and expenditures of 194 first-line budget users in central government (mainly ministries, agencies, and commissions),9 as well as the three SSFs. Combined the budget covers 95.1 percent of budgetary central government revenue.10 Expenditures of each administrative unit are presented by financing source: core budget and funds, own revenues, loans, and grants. Transfers from central government to other public sector units are also clearly disclosed in the budget. However, taxation levied by the central government and allocated by law to the PESR (excises on road fuel and taxes on motor vehicle registration which amount to 0.7 percent of GDP) are not included as central government revenue and are instead netted against equivalent transfers made to the company. In addition, taxes levied on banks and allocated to the DIF (0.2 percent of GDP) are not included in the budget. While these omissions have no net impact on the budget deficit, it means that tax revenues and expenditures are underreported in the budget by 0.9 percent of GDP. However, this may be offset by misreporting of other transactions as revenues and expenditures, as set out in principle 1.1.3.

41. The budget does not include information on the revenues and expenditures of several extra-budgetary entities. These entities, for different reasons, would be classified within central government under international reporting standards. While these entities are not central budgetary units, presenting their revenues and expenditures in the budget presentation (as is done for the SSFs), along with the taxes allocated to the PESR, would increase transparency, and see Macedonia meet the advanced level of practice as defined by the Code. The main gaps include:

• Own revenues and related expenditures of six nonfinancial regulatory bodies, which despite being legally independent and financially self-sufficient, finance themselves by means of fees rather than market prices, although they are not large (estimated to be 0.2 percent of GDP);11

• Own-revenues and related-expenditures of public health institutions (0.4 percent of GDP), although the transfers to finance their operating and capital expenses are already reflected by the budget (as part of the Health Insurance Fund and the Ministry of Health expenditure, respectively); and

9 The annual budget also contains revenues and expenditures of 194 second-line budget users (primarily universities, institutes and social or cultural centers), which are administratively subordinated to first-line users. 10 Excluding the Social Security Funds. 11 These nonfinancial regulatory bodies are the Agency for the Regulation of the Railway Sector, Energy Regulatory Commission, Civil Aviation Agency, Agency for Electronic Communications, Macedonia Postal Agency, and Agency for Audio and Audiovisual Media Services.

31

• Own-revenues and related expenditures of the PESR, DIF, MZT, MPT and some other smaller public companies, which are treated by the government as public corporations, but would be classified within the central government under GFSM 2014 standards. In total, revenues and expenditures of these corporations would account for 1.3 and 2.6 percent of GDP respectively.

42. The implications of these omissions for the reported budget aggregates is material. Combined, the above omissions understated central budget total revenues by 1.9 percent of GDP and expenditures by 3.2 percent of GDP, in consolidated terms, with an overall impact of 1.3 percent of GDP on the reported budget deficit in 2016.

43. The budget documentation presents information on financing requirements and sources, but does not include on-lending activities. The budget includes information on financing by source, including net domestic and foreign currency borrowing, along with movements in government deposits. However, loans from international financial institutions for investment purposes that are on-lent to other public sector entities are not included in total financing, although this omission does not appear to be large (see principle 1.1.3 and 1.4.3). Appropriately accounting for these transactions would have no impact on the deficit or debt aggregates, but would enhance consistency of financing and the change in debt stock.

2.1.2. Macroeconomic Forecasts (Advanced)

44. Budget documentation includes three-year forecasts for the main macroeconomic variables, their components and underlying assumptions. The ERP and, in a more summarized way the Fiscal Strategy Report (FSR), disclose comprehensive three-year projections of key economic variables. In addition, the ERP incorporates a rich and well-articulated discussion of recent economic developments, the medium-term forecasts, their components and their drivers. The assumptions underpinning the projections are disclosed in the ERP, which helps to assess the credibility of macroeconomic projections. Adding a similar table of assumptions in the FSR would enhance transparency of the macroeconomic forecasts by ensuring this information is disclosed at the time of budget deliberation.

45. Over the past decade, medium term real GDP forecast errors have been material (Figure 2.1). Since 2008, the absolute value of budget year forecasting errors for real GDP growth have averaged 1.9 percent, which is slightly above the average of European countries (of 1.3 percent). Real GDP forecasting errors remain relatively high, even after adjusted for volatility across countries, reflecting the fact that Macedonia has faced relatively lower volatility over this period than other European countries in the sample, suggesting real GDP should be relatively more predictable.12

12 The strengthening of the macroeconomic forecasting capacity of the MoF is one key objective within the PFM Reform Program 2018–21, by means of an enhancement of modelling tools -particularly regarding investment- and the use of a broadened base of indicators as forecasting inputs. The latter objective will be particularly important, given the lag between the end of the quarter and the publication of Quarterly National Accounts.

32

Figure 2.1. Medium-Term Macroeconomic Forecast Errors, 2008–17 (Percent)

a. Real GDP Forecast Accuracy – Absolute Error for Budget Year

b. Macedonia: Real GDP Growth Forecast History

c. Macedonia: Nominal GDP Growth Forecast History

Source: MoF, IMF staff estimates. Note: Volatility adjustment is average absolute forecast error divided by standard deviation of growth over the period.

0

1

2

3

4

5

6

Belg

ium

Aust

ria

Pola

nd

Fran

ce

Net

herla

nds

Spai

n

Port

ugal UK

Denm

ark

Italy

Ger

man

y

Rom

ania

Slov

enia

Slov

ak R

ep

Hung

ary

Lith

uani

a

Cypr

us

Bulg

aria

Mac

edon

ia

Czec

h Re

p

Swed

en

Finl

and

Luxe

mbo

urg

Mal

ta

Latv

ia

Esto

nia

Irela

nd

Absolute errorAbsolute error adjusted for volatility

-2

0

2

4

6

8

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

Actual

Forecasts

-2

0

2

4

6

8

10

12

14

16

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

Actual

Forecasts

33

46. Real GDP forecasts have had an optimistic bias, although the bias is less evident for nominal GDP. During 2009–12, real GDP growth forecasts were consistently stronger than outcomes, as forecasters overestimated the pace of the recovery following the global financial crisis, which has been the case in most countries. Since 2013 the forecast bias has been much more muted, although negative errors in real GDP forecasts have been significant in the past two years, partly reflecting the unanticipated consequences of political uncertainty on investment and economic growth. In contrast, inflation forecasts have tended to be higher than actuals, although there is no clear cyclical pattern to the bias. Nominal GDP growth forecasts for the budget year have not shown any significant bias in either direction and have been lower over more recent years.13

2.1.3. Medium-Term Budget Framework (Good)

47. Budget documentation includes medium-term projections of revenues, expenditures, and financing by economic category. The FSR and the ERP include three-year revenue, expenditure and financing projections, with the same institutional coverage as the annual budget, along with outcomes or estimates for the two preceding years. Medium-term expenditure plans are presented by economic classification for budgetary central government, and revenues are presented by sub-category. In addition, three-year fiscal projections are provided at the aggregate level for the consolidated general government sector. There is no presentation of medium-term expenditure plans by administrative unit or by program, as this is not required by the Budget Law. The budget does, however, include three-year projections for programs under the development budget (capital projects), which in 2018, reflected around 11 percent of total expenditures.

48. Budget deficit forecasts have been relatively accurate over the past decade, but revenue and expenditure forecasts have been systematically over estimated14 (Figure 2.2). The budget year deficit has been underestimated by less than 0.1 percent of GDP on average over the past decade, which is much lower than the average of 0.9 percent of GDP for EU countries. But, revenue and expenditures have been consistently overestimated in the budget year, by an average of 2.2 and 2.1 percent of GDP respectively.15 The government has consistently adjusted expenditure downwards during the budget year, in response to lower than anticipated revenues, in order to maintain the targeted budget balance. These adjustments have largely resulted from lower than expected capital expenditure, which has been under-executed by an average of 38 percent. Indeed, the correlation between revenue and capital expenditure errors has been noticeable, particularly during cyclical downturns (Figure 2.3).

13 Biases are measured by the median, to avoid excessive influence of large outliers. 14 Biases across in one-two and three-year ahead projections are more consistent and larger when measured by the median of errors, rather by the average. 15 Within revenues, non-tax revenues and VAT consistently show optimistic biases, while direct taxes tend to exhibit a higher cyclical correlation.

34

Figure 2.2. Average Medium-Term Fiscal Forecast Error, 2008–17 (Percent of GDP) a. Revenue Forecast Bias – Average Forecast Error

b. Expenditures Forecast Bias – Average Forecast Error

c. Budget Balance Forecast Bias – Average Forecast Error

Source: MoF, IMF staff estimates. Note: For Macedonia, data are for budgetary central government including SSFs and are missing some data for budget year +1 and budget year +2 for 2013-2015. Data for all other countries are for general government.

-8

-6

-4

-2

0

2

4

6

8

10

Bulg

aria

Cypr

usIre

land

Slov

enia

Rom

ania

Mal

ta UK

Net

herla

nds

Mac

edon

iaSw

eden

Lith

uani

aSp

ain

Italy

Pola

ndPo

rtug

alCz

ech

Rep

Finl

and

Fran

ceLa

tvia

Belg

ium

Ger

man

yAu

stria

Esto

nia

Hung

ary

Denm

ark

Luxe

mbo

urg

Slov

ak R

ep

Budget YearBY+1BY+2 Underestimate

Overestimate

-8

-6

-4

-2

0

2

4

6

8

10

Bulg

aria

Cypr

us

Rom

ania

Mac

edon

ia

Mal

ta

Net

herla

nds

Swed

en

Slov

enia UK

Czec

h Re

p

Ger

man

y

Lith

uani

a

Italy

Esto

nia

Aust

ria

Pola

nd

Finl

and

Denm

ark

Irela

nd

Latv

ia

Luxe

mbo

urg

Hung

ary

Fran

ce

Belg

ium

Port

ugal

Spai

n

Slov

ak R

ep

Budget YearBY+1BY+2

Overestimate

Underestimate

-7

-6

-5

-4

-3

-2

-1

1

2

3

Irelan

dPo

rtuga

lFin

land

Lithu

ania UK

Slove

niaSp

ainFr

ance

Belgi

um Italy

Polan

dSw

eden

Rom

ania

Malt

aSlo

vakia

Bulga

riaLa

tvia

Cypr

usHu

ngar

yAu

stria

Luxe

mbou

rgEs

tonia

Czec

h Rep

Mac

edon

iaDe

nmar

kGe

rman

yNe

ther

lands

Budget YearBY+1BY+2

Pessimistic

Optimistic

35

Figure 2.3. Budget Forecast Errors for Capital Expenditure and Revenue (Percent of GDP)

Source: MoF, IMF staff estimations.

49. Fiscal forecasts have been more accurate in recent years (Figure 2.4). Absolute revenue and expenditure forecast errors have been lower in the period 2013–17 compared to 2008–12, partly reflecting uncertainty caused by the global financial crisis in the earlier period. The same holds for deficit forecasts beyond the budget year. That said, uncertainty caused by the recent political environment has impacted on forecast accuracy, with higher forecast errors in 2016–17.

Figure 2.4. Evolution of Fiscal Forecast Errors (average absolute errors)

Revenues and expenditures (average percentage errors)

Budget Deficit (Percent of GDP)

Source: MoF, IMF staff estimates. Note: Data are for budgetary central government including SSFs.

2008

2009

2010

2011

2012

20132014 2015

2016

2017

-30

-25

-20

-15

-10

-5

0

-5 -4 -3 -2 -1 0 1 2

CAPE

X er

rors

Revenue errors

0

2

4

6

8

10

12

14

2008-2012 2013-2017 2008-2012 2013-2017

Budget Year

BY+1

BY+2

Revenue Expenditure

0

1

2

3

4

5

6

2008-2012 2013-2017

Budget YearBy+1BY+2

36

50. The Government intends to strengthen the medium-term budget framework (MTBF) as part of its broader PFM reform strategy. The 2018–21 PFM Reform Program aims at improving fiscal forecasting tools and developing a comprehensive MTBF that links sectors, programs, and projects to ensure spending allocations are aligned with policy objectives and make planning more stable. Against this framework, the government, with the support of the Finnish Twinning Project has developed a manual for MTBF preparation, which describes in detail the formulation of budget requests by administrative units, programs and sub-programs. Implementation will follow the adoption of a comprehensive program structure for the budget. The new budget structure would present expenditure plans during the budget year and the two following years by administrative unit and, within each unit, by program and sub-program. Successful implementation of these plans, would see Macedonia meet the advanced level of practice as defined by the Code.

2.1.4. Investment Projects (Basic)

51. The government provides substantial information on public investment projects, but does not regularly disclose the total multi-annual financial obligations for them. The annual budget provides information on the budget allocation for capital expenses of investment projects for the three-year projection period and provides a summary for each project or sub-program included under the development budget. But it does not include the total life-cycle cost for those projects, amounts already spent, and any deviations in project costings from the previous budget. The FSR and the ERP also provide detailed information on public investment priorities, but do not comprehensively report total project costs. The National Investment Council published a list of national priority projects in 2015, which includes some information on the projects total cost, financing source, estimated construction period, and stage of development. But, the list does not allocate costs over the life of the project and it provides rather limited information on some of the sectors.

52. Cost-benefit analysis (CBA) for major investment projects are not always required prior to their approval and are not published. For investment projects financed by the EU, international financial institutions and international development banks, cost-benefit analyses are generally undertaken according to the individual donors’ approved methodologies. Although most of the major infrastructure projects are subject to this procedure as a pre-condition for external financing, there is no legal requirement to conduct and publish CBA for investment projects for those financed by either external bilateral loans or the budget. The new Manual for Medium-Term Budgeting underscores the need for CBA for new project proposals, particularly the largest ones, without limiting this obligation according to the source of financing. However, the Manual lacks legal authority and no methodology has so far been developed to guide CBA application.

37

53. Major investment projects are subject to open and competitive bidding, but with some exceptions. The Public Procurement Law requires competitive tendering of those contracts of supplies, services, and public works16 above 500 euros (on a monthly basis). The law sets out some exceptions (Articles 7-9), related to defense, security, and public contracts financed by international organizations or third countries provided they have their own awarding systems. Most of these are in conformity with EU procurement directive, except for the exemption related to national security. In 2016,17 76 percent of the contract value of tenders was awarded through open procedure,18 while only 7 percent pertained to restricted or negotiated procedures without prior publication of a contract notice and 9 percent was conducted through simplified procedures. However, the number of bidders in the open procedure has been low (around 3.5 on average), which can impact on its effectiveness. The government is planning amendments to the procurement law to address its main shortcomings and improve transparency and efficiency of the system, by bringing it into compliance with the EU 2014 procurement directive.

54. The Public Procurement Bureau also publishes information on tenders and their outcomes, as well as administrative decisions on the appeals against tenders. The electronic platform of the Public Procurement Bureau transparently discloses information over the life of contracts. Complaints about procedures and tender outcomes can be addressed to the State Appeals Commission, that decides on the legality of the actions of the authorities and awarding decisions and its decisions are published in the Official Gazette of the Republic of Macedonia and the Bureau’s website. In 2016, out of 623 appeals received, around 250 were accepted by the Commission.

55. General government investment in Macedonia is relatively high in comparison with peer countries. General government investment (including those of the PESR and Macedonia’s Railway Infrastructure) stood at around 5.2 percent of GDP in 2016 (Figure 2.5) and is projected to increase further over the medium term.19 Total public investment is higher, at 5.8 percent of GDP in 2016. That said, planned large-scale ramp up in public investment has proved difficult in previous years, with capital expenditures typically under-executed. On average, over the past 10 years, capital expenditure has been 38 percent lower than budgeted. The government has outlined its intention to strengthen public investment management as part of its PFM reform program.

16 Procurement of concessions and PPPs is regulated by the Law on Concessions and PPPs. 17 See Annual Report 2016 of the Public Procurement Bureau. 18 Restricted contracts are those where the contracting authority may select -and exclude- bidders out of received tenders, while negotiated procedures imply consultations of the contracting authority with one or more of the selected bidders to agree the terms of the contract. By contrast, in open procedures all participating bidders compete on an equal footing. 19 Government investment in the authorities own fiscal reports is 2.9 percent of GDP, but excludes investment by extrabudgetary units

38

Figure 2.5. General Government Investment in Emerging Europe and Central Asia, 2016 (Percent of GDP)

Source: Macedonian authorities, IMF.

2.2. Orderliness 2.2.1. Fiscal Legislation (Advanced)

56. The Budget Law defines the timeframe for budget preparation, the key contents of the executive proposal and the powers of the National Assembly to amend the budget. The Budget law defines a clear timetable for budget preparation. It requires that the government adopt a three-year fiscal strategy by end-May, which forms the basis for budget preparation, and submits its draft budget to the National Assembly before November 15, to be adopted by the end of the year. The Budget Law specifies the content the Executive Budget Proposal (article 29)20 and fiscal strategy (Article 17). Further, the Rules and Procedures of the National Assembly (published on its website) detail the procedure for its discussion (Articles 179 to 181). In particular, the National Assembly is entitled to reject the Draft Budget and propose amendments, which will oblige the government to present a revised Draft Budget with explanatory notes.21 The Budget Law does not currently require the fiscal strategy to be either submitted or debated by the National Assembly, although it is provided for information purposes. Requiring that the FSR

20 The proposal shall comprise a general part with the basic fiscal aggregates, a special part with appropriations by budget user, programs and sub-programs, and a development part. 21 The Budget Law adds some limits to the powers of the Parliament. During the discussion of budget in the National Assembly, increases in some appropriations shall be offset by a corresponding reduction in others. In addition, the reserve in the Draft Budget shall not be reduced to increase other appropriations.

0

1

2

3

4

5

6

7

8

9

Kyrg

yz R

epub

lic

Koso

vo

Turk

ey

Mac

edon

ia

Alba

nia

Mol

dova

Rom

ania

Russ

ia

Arm

enia

Pola

nd

Kaza

khst

an

Uzb

ekist

an

Geor

gia

Lith

uani

a

PESR and MZT

Budgetary General Government

39

be submitted and debated in parliament would strengthen legislative scrutiny of fiscal policy and performance.22

2.2.2. Timeliness of Budget Documents (Good)

57. The draft budget has generally been submitted to the National Assembly at least two months before the end of the fiscal year. The Budget Law requires that budget be submitted to the National Assembly by November 15, but submissions during the past five years have typically taken place earlier, in mid-September (Table 2.2). The exception was 2018, where the budget was submitted in early November. The National Assembly has consistently approved the budget in December, prior to the start of the next fiscal year. Although the Budget Law requires that the FSR be published by end May, this has not been the case in each of the past three years, where the FSRs were published in October or later, with the delays at least in part attributable to uncertainties around the political environment. Publishing the FSR, at or around the same time as the budget, dilutes the depth of scrutiny that can be given to each of these documents.

Table 2.2. Macedonia: Dates of Budget Submission and Approval

2.3. Policy Orientation 2.3.1. Fiscal Policy Objectives (Basic)

58. The FSR specifies a numerical limit on public debt, complemented by rolling medium-term deficit and debt targets. The Law of Public Debt requires that fiscal authorities set a limit for public debt as a share of GDP, with public debt defined as general government debt securities and loans and guaranteed debt. Since the 2017 FSR, this limit has been established at 60 percent of GDP, and is accompanied by other rolling medium term (over a three-year time horizon) and short-term (within one year) objectives, in terms of the composition of public debt.23 The 60 percent of GDP debt ceiling can be considered a precise objective, but not time-bound. In addition, every Fiscal Strategy sets a path of rolling revenue, expenditure, deficit, and debt projections for general government over the same three-year horizon. However, these are indicative and non-binding, and tend to be revised from year to year (Table 2.3).

22 The draft Strategic Program 2018-21 of the National Assembly includes among its objectives analysis and discussion of the fiscal strategy. 23 Medium-term objectives refer to a cap for the level of guaranteed debt (in terms of GDP), a floor for euro-denominated debt within FX-denominated debt, a minimum threshold for fixed-interest rate debt out of total debt. Short-term objectives are established in terms of net public debt borrowing, net guaranteed borrowing, a minimum threshold for average time to maturity and another minimum threshold for average time to re-fixing.

2014 Budget 2015 Budget 2016 Budget 2017 Budget 2018 BudgetSubmission to Parliament 20-Sep-13 13-Sep-14 23-Sep-15 19-Sep-16 9-Nov-17Approval by Parliament 20-Dec-13 16-Oct-14 20-Nov-15 17-Oct-16 28-Dec-17

40

Table 2.3. Macedonia: Fiscal Targets and Objectives Compared to Outturns (Percent of GDP)

Source: Ministry of Finance and Fiscal Strategy Reports. Note: Deficit and debt are for the general government. The 2015 FSR did not include a projection for general government debt, only for the aggregate of general government and government guaranteed debt.

59. The government provides the necessary information to assess compliance with the debt limit, but does not analyze its performance against the fiscal strategy. The FSR sets out three-year projections for public debt, while the Annual Report on Public Debt Management discusses outcomes for the level of public debt and its composition. But, neither document provides a specific discussion on the rationale behind the fiscal objectives, consistency of projections against the fiscal objective, or extent of past compliance.

60. The government has committed to strengthening the fiscal rules framework. The 2018–21 PFM Reform Program aims to adopt fiscal rules legislation by 2019. To ensure transparency, the government should clearly articulate the rationale for the selected fiscal rules in its FSR and include a discussion on the compliance of medium-term fiscal forecasts, and outcomes (once the rule is established) with those objectives. The credibility of a government’s fiscal objectives can also be enhanced by ex ante specification of situations under which the government could deviate from its stated fiscal objectives.

2.3.2. Performance Information (Basic)

61. The budget presents program information for each administrative unit, as well as a breakdown of expenditures allocated to each program by economic classification. But, there is no link between most of these programs and their policy objectives, and an absence of performance indicators for each program. The budget provides information on 11 government programs, which relate to activities across administrative units, along with a description of their main qualitative objectives and an identification of managing ministries.24 The budget also defines output indicators for administrative units, but it is difficult to link these targets to the different programs they are responsible for administering and there is no reporting on performance against them.25 Explaining the policy objectives of all programs, and assigning clear

24 These priority programs are: decentralization, poverty reduction, defense and security, strengthening the right of law, economic development, reforms in the economic administration, integration in the EU, information and communication technologies, regional development, environmental protection and investment in education. 25 For example, the Ministry of Labor and Social Policy includes as one of its output indicators, the number of children receiving benefits, but this is not linked to a specific social benefit program.

Deficit Debt Deficit Debt Deficit DebtFSR 2015 -3.3 n.p -3.2 37.3 -2.8 38.7FSR 2016 -3.2 39.4 -2.9 40.7FSR 2017 -3.0 42.7Actual -3.4 38.1 -2.7 39.6 -2.9 39.3

2015 2016 2017

41

output targets to each of them, rather than to institutions, would facilitate greater policy analysis and provide better information on the effectiveness of spending in meeting those objectives.

62. Work is underway to develop a comprehensive structure of programs and sub-programs, with clear output indicators. A Manual on Program Budgeting has been developed, which proposes a new programmatic structure based on 14 policy areas and 49 programs each of them assigned to one or more administrative units. Once fully implemented, the budget would detail the objectives of each of the programs, and an annex (or a separate publication of the MoF) would include tables linking each sub-program with performance indicators (in terms of realized values for past years and objectives for the next three years). Implementation will first require a review of the organizational structure of administrative units.

63. The government is also introducing a gender perspective in budgeting and published for the first time in 2018 gender budget statements for certain ministries. In 2012, the government adopted a Gender Equality Strategy (2013–20), which required the inclusion of a gender perspective in the policy planning and budgeting process. Pilots have been underway in various ministries, with each ministry selecting a pilot program in 2017 to analyze from a gender perspective, including through the development of gender indicators to monitor its impact. In 2018, each of the pilot ministries presented a gender initiative within one of their sector programs and prepared gender budget statements.

2.3.3. Public Participation (Good)

64. The Government published a citizen’s guide to the budget for the first time in 2017, for the 2018 Budget. The citizens budget provides a summary explanation of macroeconomic projections underlying the budget, the main fiscal aggregates, the composition of budget revenues and expenditures, and summary of the main policy priorities in the budget. The information contained in the citizen’s budget could be enhanced by linking key spending initiatives to the government’s policy objectives and by providing greater detail on how the main policy initiatives contained in the budget impact on the typical citizen, or different groups of individuals. Line ministries have also made commendable efforts to convey to citizens how their budget allocations will be spent.

65. Public participation in budget deliberations started in 2017 and further steps could be made in this direction. A first step to provide the public with a formal voice in the budget process was made in 2018, with stakeholders from civil society groups, business, labor market unions, and academia, invited to participate in a presentation and discussion of the draft budget. There is scope to enhance opportunities for public participation in the budget process by opening the forum to a wider range of stakeholders at an earlier stage of budget preparation. To allow for more open and inclusive participation, the government could introduce a formal call for

42

budget submissions from the public to help inform policy priorities, or introduce a public participation component in parliamentary committee hearings.26

2.4. Credibility 2.4.1. Independent Evaluation (Not met)

66. Budget documentation does not include a comparison of fiscal projections with those of external forecasters and there is no independent evaluation of them. The FSR and the ERP include a comparison of medium-term real GDP and inflation forecasts with other forecasters (IMF, WB, EBRD, EC, Vienna Institute, and NBRM). But, no comparison is provided for the fiscal forecasts. Given the historical magnitude of deviations in revenue and expenditure outturns from forecasts, including a comparison of forecasts for the main fiscal aggregates would assist users of fiscal information in judging the realism of the government’s budget forecasts. There are currently no independent fiscal agencies in Macedonia that produce their own fiscal and economic forecasts or assess the government’s forecasts and fiscal performance, though the government committed to establishing a fiscal council in its 2018–21 PFM Reform Program.

2.4.2. Supplementary budget (Good)

67. The Budget Law requires parliamentary approval for changes to total budget expenditure, but allows for moderate changes to its composition. The Budget Law requires legislative approval for any changes to the total budget allocation (Article 36) or to allocations across budget users (Article 33(7)). However, the Budget Law allows for individual budget users to reallocate within their own budgets, without legislative approval, provided this does not reduce expenditures within an economic classification by more than 20 percent or increase salaries by more than 10 percent (Article 33(1-6)). Reallocations within budget users of up to MKD 5 million require approval from the MoF, while amounts above this require approval of the Council of the government. In addition, Article 11(3) establishes that, in the event of a national crisis, appropriations of the contingency reserve can be increased (up to its cap of 3 percent of total expenditure) by reallocating spending from elsewhere in the budget provided these do not result in reductions of an individual budget user appropriations of more than 5 percent.

68. Supplementary budgets, along with regular under-execution of approved expenditures, have helped restrain spending in response to lower than expected revenues. During the past decade, supplementary budgets were approved on nine occasions, with five of these resulting in reduced budget appropriations (Figure 2.6). Excluding 2008, supplementary budgets have resulted in reduced expenditures of 1.5 percent, on average, compared to the

26South Africa parliamentary committees are required to invite civil society and experts when the budget is tabled, and the public can make written submissions and participate in hearings. A parliamentary budget office is in charge of collecting input from civil society and presenting it to Parliament. In Korea, the Special Budget Committee invites experts to debate in a broadcasted hearing on issues related the economic and tax forecasts, fiscal stance and expenditure programs. In Australia, there is an open call for budget submissions inviting all stakeholders to set out their priorities or proposing specific measures for consideration as part of the budget.

43

original budget, although there have been occasions where the adjustment has been much higher (such as in 2009). For the period 2009–17, the spending outturn was, on average, 4.8 percent lower than the spending approved in supplementary budgets, and 6.3 percent lower than originally budgeted.

Figure 2.6. Components of Deviation in Outcome from Approved Budget (Percent of Approved Budget)

Source: MoF and staff.

2.4.3. Forecast Reconciliation (Not Met)

69. The budget documentation does not explain changes between successive forecasts or medium-term plans. The ERP includes comparison of medium-term forecasts for real GDP growth, the budget deficit and debt between the current and the previous fiscal strategy. But, it does not compare changes in forecasts for the key budget components (revenues and expenditures). The ERP quantifies the budgetary impact of the main structural reforms of the government over the next three years, but there is not a full discussion of new initiatives with a budgetary impact,27 or how they interact with macroeconomic developments to explain departures from previous fiscal objectives. Including a more detailed comparison of macroeconomic and fiscal variables across vintages of fiscal strategies in the FSR as well as ERP, along with an explanation of the main reasons for revisions would enhance the credibility of the MTBF and assist with assessments of compliance with the fiscal strategy.

70. Revisions to expenditure plans have been significant over the past decade, although they have been smaller in more recent years (Figure 2.7). Between successive fiscal

27 The government intends to expand the elaboration on the budgetary impact of some additional initiatives in the next Fiscal Strategy Report.

-20

-15

-10

-5

0

5

10

15

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Average

ExecutionSupplementary BudgetOutturn

44

plans, absolute revisions to three-year ahead28 expenditure plans averaged 7.4 percent over the past ten years, and revisions to the second-year ahead year averaged 6.2 percent. Revisions were predominantly larger prior to 2011, partly due to the uncertainty surrounding revenue forecasts in the context of the global financial crisis. Since then, the average size of revisions has decreased substantially but are still material.

Figure 2.7. Revisions to Medium-term Plans, 2008–17 (Percent of total expenditure)

Source: IMF staff estimates. Note: Available data for 2013–14 do not allow the calculation of revisions. The year refers to the year for which the forecast was made. The columns show the revisions to expenditure plans made for each year over time. For example, for 2010, it shows the change in the 2010 forecast made in the 2008 fiscal strategy report (published in 2007) compared to the forecast made in the 2009 fiscal strategy report (BY+2 to BY+1) and again, the forecast for 2010 made in the 2009 fiscal strategy report, compared to the forecast made in the 2010 fiscal strategy report (BY+1 to BY).

2.5. Recommendations 71. Fiscal transparency in budgeting and forecasting follows good practices in several areas. The assessment against the Code, summarized in Table 2.4, shows budget documentation includes medium-term macroeconomic and fiscal forecasts, a draft budget is prepared in accordance with the provisions of the Budget Law and presented to the legislature around two months before the end of the year, with its core components summarized for the public in a Citizens Guide, and that legislative approval is required for any changes to aggregate expenditures.

28 For example, in the fiscal plan 2015–17, 2017 is the third year ahead (BY+2, in the notation we are using in this report). The revision would firstly quantify the percentage change in expenditure plans for 2017 between the fiscal strategy 2015–17 and 2016–18, and then between the strategy 2016–18 and 2017–19.

-15

-10

-5

0

5

10

15

20

25

2008 2009 2010 2011 2012 2015 2016 2017

BY+2 to BY+1BY+1 to Budget Year

Spending revised upwards

Spending revised downwards

45

72. However, there is scope to enhance the comprehensiveness and credibility of the budget. Not all central government revenues and expenditures are reported in the budget. While numerical fiscal rules exist, they are not well integrated with medium-term budget planning, and differences between successive medium-term plans are not adequately explained. There is also scope to enhance transparency around public investment, including by strengthening reporting on their fiscal costs and processes for their selection and appraisal.

73. Based on the above assessment, the evaluation highlights the following priorities for improving transparency of fiscal forecasts and budgets:

• Recommendation 2.1: Improve budget comprehensiveness. Include in the budget, the taxes allocated to the PESR and DIF, and associated transfers, and present information on the own revenues and expenditures of all extrabudgetary funds and include these in summary tables of consolidated budget revenue, expenditure, and balance.

• Recommendation 2.2: Strengthen the credibility and transparency of the MTBF.

a. Publish multi-annual expenditure projections by administrative unit.

b. Explain reasons for possible deviations in key fiscal aggregates between successive fiscal forecasts, including by reporting the extent that these are due to policy decisions of the government or a result of economic and other developments.

c. Include comparison of fiscal forecasts (revenues, expenditures, deficit and debt) to those of independent forecasters.

• Recommendation 2.3: Improve the policy orientation of the budget. Include in the FSR a clear statement on the government’s fiscal objectives and assessment of whether medium-term fiscal plans, as well as past outcomes, are consistent with them and regularly report on the performance of major programs against clearly defined indicators.

• Recommendation 2.4: Enhance the transparency of public investment projects. Publish the total costs of major investment projects in addition to their multi-annual allocation, and, develop and publish methodological guidelines for the appraisal of large projects, along with their assessments.

46

Table 2.4. Macedonia: Summary Evaluation: Fiscal Forecasting and Budgeting Principle Assessment Importance Rec

2.1.1 Budget Unity Not met: The budget reports most revenues and spending of central government units on a gross basis, but excludes taxes allocated to the PESR and DIF.

High: The exclusion of taxes redirected to the PESR and DIF understates central government revenues and expenditures by 0.9 percent of GDP.

2.1

2.1.2 Macroeconomic Forecasts

Advanced: Budget documentation includes medium-term macroeconomic projections, their components and underlying assumptions.

Medium: Real GDP forecasts for the budget year have been, on average, overestimated by 1.9 percent (in absolute terms), which is slightly more than the EU average.

2.1.3 Medium-Term Budget Framework

Good: Budget documentation includes medium-term fiscal projections of revenues, expenditures and financing by economic category.

Medium: Revenue and expenditures have been overestimated by an average of 2.2 and 2.1 percent of GDP in the budget year, but deficit forecast errors have been low.

2.2

2.1.4 Investment Projects

Basic: All major projects are subject to open and competitive tender, but total costs of individual projects are not regularly disclosed, and cost-benefit analyses are not published.

High: Public investment was 5.8 percent of GDP in 2016 and there has been large under-execution (38 percent annual average).

2.4

2.2.1 Fiscal Legislation

Advanced: The Budget Law defines the timetable for budget formulation, its content and the legislature’s powers to amend it.

Low: While the core content of budget documents is included in the legislation, there is scope to expand the requirements further.

2.2.2 Timeliness of Budget Documents

Good: The budget is regularly submitted 2 months before the start of the financial year and approved by Parliament before the start of the year.

Medium: FSR has been published with delay, limiting its effectiveness and the capacity of the legislature and public to scrutinize information.

2.3.1 Fiscal Policy Objectives

Basic: The FSR defines a numerical limit for public debt, but this is not time-bound.

Medium: There is limited assessment of fiscal policy in budget documents, but fiscal targets are generally met.

2.3

2.3.2 Performance Information

Basic: Budget documentation includes input allocation by policy areas.

Medium: Lack of clear linkages between policy objectives defined for each institution and programs.

2.3

2.3.3 Public Participation

Good: A Citizen’s Budget is published, and major stakeholders are consulted as part of the budget deliberation process.

Low: Budget documentation is complex, comprising more than 600 pages, but is summarized and discussed with stakeholders.

2.4.1 Independent Evaluation

Not met: Fiscal forecasts are not compared to those of external institutions and there is no independent evaluation of them.

Medium: Fiscal forecasts present systematic errors, both on revenue and expenditure.

2.2

2.4.2 Supplementary Budget

Good: Changes to the level of budget expenditure require parliamentary approval.

Low. Supplementary budgets have reduced expenditure by 1.5 percent a year on average, since 2009.

2.4.3 Forecast Reconciliation

Not met: Differences in revenue and expenditures from previous medium-term forecasts are not explained.

High: Absolute revisions to expenditure plans have average 7 percent a year from 2008.

2.2

47

III. FISCAL RISKS 74. Governments should disclose, analyze, and manage risks to the public finances and ensure effective coordination of fiscal decision-making across the public sector. This chapter assesses the quality of Macedonia’s fiscal risk analysis, management, and reporting practices against the standards set by three dimensions of the IMF’s Fiscal Transparency Code:

• General arrangements for the disclosure and analysis of fiscal risks;

• The reporting and management of risks arising from specific sources, such as government guarantees, public-private partnerships, and the financial sector; and

• Coordination of fiscal decision-making between central government, local governments, and PCs.

75. Disclosure and management of fiscal risks in Macedonia has improved considerably over recent years.29 The government discloses and assesses many of the fiscal risks it faces, including from macroeconomic shocks, the financial sector, and long-term risks associated with pension and health care expenditure as part of its reporting obligations as an EU candidate country. Table 3.1 lists the various reports published by the government that serves as the basis for fiscal risk analysis and management in Macedonia.

Table 3.1. Macedonia: Reports Related to Fiscal Risks

Report Related Risks and Issues Author

Economic Reform Program (ERP)

Macroeconomic risks (including scenario analysis), debt portfolio risks (including sensitivity analysis), external debt sustainability, long term fiscal sustainability, guarantees

MoF, NBRM, PPB, Ministries.

Annual Report on Public Debt Management of the Republic of Macedonia

Debt management and Guaranteed Public Debt. MoF

Financial Stability Report for the Republic of Macedonia Financial Sector Stability NBRM

Deposit Insurance Fund Annual Report

Explicit exposure to the financial sector, asset portfolio risks (including sensitivity analysis)

Deposit Insurance Fund

Annual Report on the Budget Contingency Reserve MoF

Fiscal Strategy Report (FSR) Macroeconomic and specific fiscal risks (including arrears and guarantees), debt portfolio risks. MoF

29 At the time of Macedonia’s Fiscal Transparency Report on the Observance of Standards and Codes (ROSC) in 2006, fiscal risk disclosure was largely confined to periodic publication of public debt and guaranteed debt.

48

3.1. Disclosure and Analysis 3.1.1. Macroeconomic risks (Good)

76. Macedonia publishes detailed analysis of macroeconomic risks and their implications for public finances. The ERP examines the impact of four pessimistic macroeconomic scenarios on the budget deficit. Two scenarios are devised using historical data to create plausible, but pessimistic, scenarios while the remaining two consider an arbitrary negative shock to revenue and expenditure. The ERP also includes analysis of the sensitivity of public debt to both interest rates and exchange rates. The FSR also contains a detailed discussion of alternative macroeconomic scenarios, but stops short of presenting their implications for the main fiscal aggregates. Bringing forward the analysis already contained in the ERP to the FSR, would allow for macro-fiscal risks to be considered at the time of fiscal policymaking. An advanced practice would take the existing analysis further by presenting probabilistic forecasts of the main fiscal aggregates.

77. The Macedonian economy has been subject to some volatility over recent years creating uncertainty for public finances. Nominal GDP and revenue growth volatility over the period 2000–17 has been higher than other European countries, on average, but volatility has been similar to that of several other southeast European countries (Figure 3.1). This volatility, combined with the relatively short maturity of government borrowing, could make fiscal risks more difficult to manage.

Figure 3.1. Volatility of Nominal GDP and Revenue Growth, 2000–17 (Percent)

Source: IMF WEO database, April 2018. Note: Volatility is measured as the standard deviation of the annual growth rate in the period 2000-2017.

CYP

CZE

EST

FRA

DEU

GRCHUN

IRL

ITA

LAT

LTH

LUXMAL

POLPRT

SLKSLV

ESP

SWE

UK

ALB

BIH

BULCRO

MKD

0

2

4

6

8

10

12

14

0 2 4 6 8 10 12 14

Reve

nue

grow

th

Nominal GDP growth

49

3.1.2. Specific Fiscal Risks (Not Met)

78. Macedonia publishes information on several of its fiscal risks in various reports, but no report summarizes the range of risks that the public finances are exposed to. The ERP discloses and quantifies exposures related to guarantees and public debt, and these are also discussed in annual debt management reports. In addition, the FSR discusses risks related to payment arrears. The Financial Stability Report analyses the risks emanating from the financial sector while the annual report of the DIF describes the risks related to guaranteed bank deposits. But, no report offers a comprehensive summary of specific fiscal risk, their magnitude, and possible implications for public finances. Publishing a summary statement of fiscal risks would help ensure the government has an aggregate picture of its risk exposures and their potential interactions.

79. Specific fiscal risks in Macedonia are sizeable, with the maximum exposure estimated at around 49 percent of GDP (Table 3.2). However, this is lower than many other countries for which data is available. The main exposure arises from the deposit guarantee which is equivalent to around 26 percent of GDP, and public corporations,30 most of whose borrowing is backed by government guarantee. Additional fiscal risks arise from natural disasters and legal cases.31 The estimates of specific fiscal risks presented in Table 3.2 represent an upper limit—many of these are assessed to be low, and even were fiscal risks to materialize from these sources, the fiscal impact could be lower.

Table 3.2. Macedonia: Selected Specific Fiscal Risks

Specific Fiscal Risk Magnitude

(Percent of GDP) Reporting

Explicit Risk Exposure Net exposure to financial sector 26.4 NBRM and DIF annual report Guaranteed public debt 8.8 Debt management report Implicit Risk Exposure Public corporation non-equity liabilities (excluding guaranteed debt)

6.6 Not reported (IMF estimate)

Public-private partnerships (total investment)

6.8 Not reported (IMF estimate)

Contingent events Natural disasters 0.3 Not reported (Global Assessment Report) Legal cases 0.03 Annual Report on Budget Execution

30 In this chapter, public corporations refers to public enterprises that are currently not included in general government in the MoF’s fiscal reports. This differs from the approach presented in Chapter I, which classifies several PCs as general government entities in line with international statistical standards. Should the authorities bring those enterprises into the general government in their fiscal reports, this will increase reported general government debt, and reduce both the explicit and implicit fiscal risk exposures related to them. 31 There are a large number of legal cases (around 12,000 enforcement orders annually), although the fiscal costs against core budget users have been low—averaging 0.03 percent of GDP annually. This figure does not include legal cases against other public institutions.

50

3.1.3. Long-term sustainability of public finances (Basic)

80. The government publishes a regular assessment of long-term fiscal sustainability. The ERP includes projections of total revenue, total expenditure, and by extension the budget deficit up to 2060. The analysis sets out the key macroeconomic and demographic assumptions which underpin the forecasts of health expenditure, pension expenditures and contributions.32 The analysis does not include the consequences on public debt. Furthermore, given long-term projections are highly uncertain and sensitive to the underlying assumptions, the analysis could be expanded further by including multiple scenarios with varying demographic and macroeconomic assumptions. This would improve understanding of the risks around long-term public finances and can be useful in understanding whether policy changes are required to ensure sustainability. One simple approach would be to use the UN’s three population forecasts which include a medium (central projection), a high (assumes a higher fertility rate), and a low scenario (assumes a lower fertility rate).

81. While Macedonia currently has relatively favorable demographics, its population is expected to age fairly quickly over the next few decades. The old-age dependency ratio in 2015 at less than 20—less than one person aged 65 or older for every five working age people—is amongst the lowest in comparator countries. However, this ratio is expected to rise relatively quickly, to 42 (i.e., more than two old-aged people for every five working age people) by 2050 (Figure 3.2). Past reforms to introduce mandatory private pension schemes and eliminate general early retirement will help mitigate some of the impacts of ageing on public finances and further reforms are planned. Nonetheless, World Bank analysis suggests that despite these initiatives, the ageing population is likely to increase the pension deficit by around 10 percent of GDP between now and 2030 under a conservative scenario.33 Health care costs are also expected to rise, by around 1 percent of GDP between now and 2050 (according to the authorities estimates), but are still expected to remain significantly below current health care expenditure in the EU.34

32 Forecasts in the ERP are based on MoF calculations and differ from the estimates of pension expenditures and contributions by the Ministry of Labor and Social Policy, which are based on actuarial projections. 33 Source: World Bank, FYR Macedonia Public Expenditure Review Fiscal Policy for Growth, 2015. 34 World Bank data on domestic general government health expenditure data shows the average across the EU in 2015 was 7.8 percent of GDP, higher than the forecast for Macedonia in 2050.

51

Figure 3.2. Old-Age Dependency Ratios (Age 65+ to 15-64 population)

Source: United Nations, World Population Prospects: The 2017 Revision.

3.2. Fiscal Risk Management 3.2.1. Budgetary contingencies (Good)

82. The budget includes two contingency reserves for unexpected events with clear access criteria, but there is limited in-year reporting on its use. Article 11 of the Budget Law sets out that the two reserves (the permanent reserve and current reserve) can only be used for unforeseen circumstances. Furthermore, the permanent reserve can only be used for dealing with natural disasters, epidemics, and environmental catastrophes. The damages associated with these events are determined independently by the Commission for Estimating and Determining the Amount of Damage from Natural Disasters. Article 11 also caps the total size of the combined reserve at 3 percent of current expenditures of the Budget. The government publishes information on the overall execution of the reserves in the Annual Report on the Budget, but with no specifics on the purposes for which it was drawn. The MoF prepares more detailed information on reserves use for internal purposes, the publication of which would see Macedonia meet advanced practice.

83. The size and use of contingency reserves, outside of dealing with large-scale natural disasters, has been relatively modest. On average, the contingency reserve included in the original budget has been much lower than the legal limit, and use of reserves has averaged even less, at around 0.1 percent of expenditure (Figure 3.3). The exception was 2015 and 2016 (i.e., the years with significant flooding), when reserve use was significantly higher at around 0.4 percent of total expenditure.

0

10

20

30

40

50

60

70

80

Swed

en UK

Fran

ce

Serb

ia

Belg

ium

Bulg

aria

Cana

da

Turk

ey

Mon

tene

gro

Ger

man

y

Rom

ania

Cypr

us

Croa

tia

Mac

edon

ia

Aust

ria

Bosn

ia

Czec

h Re

publ

ic

Alba

nia

Slov

akia

Ital

y

Pola

nd

Slov

enia

Port

ugal

Gre

ece

Spai

n

Increase 2015-2050

2050

52

Figure 3.3. Use of the Contingency Reserves (Percent of Expenditure)

Source: Budget documents and MoF data.

3.2.2. Management of assets and liabilities (Basic)

84. Government borrowing is authorized by law, and the risks surrounding government debt and some financial assets are analyzed and disclosed. The FSR sets limits on new borrowing and a limit for public debt at 60 percent of GDP.35 Information on the stock of public debt is published quarterly, within a month of the end of each quarter. The annual report on debt management discusses exchange rate, interest rate, and rollover risks associated with general government debt, as well as strategies to manage these, while the ERP reports the sensitivity of public debt to interest rates and to the euro exchange rate (to which the Denar is pegged). However, the analysis does not extend to assessment of market exposures associated with guaranteed debt, some of which are provided on foreign currency borrowing.

85. Risks to public debt primarily relate to the maturity of the debt portfolio and interest rates. The government reported its stock of general government debt at 39.1 percent of GDP at end-2016.36 Of this amount, around 60 percent is external debt. While this exposes the debt portfolio to exchange rate risk, most foreign currency debt (about 90 percent) is denominated in euros, the risks around which are mitigated by the exchange-rate regime. Key risks relate to the maturity of the debt portfolio and exposure to interest rates. The domestic debt portfolio has an average time to maturity and average time to re-fixing of around three

35 Public debt is defined in the Public Debt Law as consisting of general government debt and guaranteed debt of public enterprises. 36 This does not include the debt of those publicly-owned enterprises, which are treated as part of general government in other sections of this report.

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

2011 2012 2013 2014 2015 2016 Averagepre-2015

Average

Budget Additional allocation in Supplementary Budget Used

53

years, while for external debt these are around 5.5 and 4.5 years respectively (Figure 3.4 and 3.5).37

Figure 3.4. Central Government Debt Maturity Profile (Years)

Average time to maturity Average time to re-fixing

Source: Debt Management Report.

Source: Debt Management Report.

86. The government manages these risks through specific limits on exposures and publishes an annual debt management strategy as part of the FSR. The debt management strategy defines limits to manage risk exposures. These include: minimum limits for average time to maturity and average time to re-fixing of three years and two years respectively; a minimum threshold for euro-denominated debt in the foreign currency-denominated portfolio of 80 percent; and, a minimum threshold for fixed interest debt of 50 percent. The government is also pursuing a strategy to lengthen the maturity profile of its debt to help counteract this risk.

87. Risks relating to non-debt liabilities, largely associated with arrears, are discussed in the FSR. There is not currently a robust system for monitoring arrears, although the MoF conducted and published in May 2017 a survey of the stock of arrears across all public sector entities, which were estimated to be around 3.6 percent of GDP. However, these are unaudited claims based on information submitted by general budget users and other public institutions, which include intra-public sector payments and possibly amounts for invoices received, but that had not fallen past due. Active central monitoring and reporting of arrears by age will be aided by a software system being developed to more accurately record arrears coupled with sanctions introduced through the Fiscal Discipline Law.

37 The average time to maturity measures the weighted average time to maturity of all the principal payments in the portfolio. The average time to re-fixing is a measure of weighted average time until all the principal payments in the debt portfolio become subject to a new interest rate.

0

1

2

3

4

5

6

2012 2013 2014 2015 2016

Domestic debt External debtTotal government Current Limit

0

1

2

3

4

5

6

2012 2013 2014 2015 2016

Domestic debt External debtTotal government Current Limit

54

88. There is limited discussion of the risks around the government’s financial assets. At the end of 2016, the general government’s financial assets were estimated to be around 51 percent of GDP (based on the current sectoral classification adopted by the authorities in their fiscal reports). These mainly comprised: equity in wholly owned PCs (37 percent of GDP); minority shares in joint stock companies (1 percent of GDP); and currency and deposits (6 percent of GDP). In addition, the central government has loans which are on lent to municipalities and public enterprises (1.5 percent of GDP). The DIF has a small asset portfolio (2.3 percent of GDP), but its investments are in public debt securities, which are consolidated out in the calculation of general government assets. Other than for the DIF, the government does not publish information on its assets, or discuss risks surrounding them.

• Equity in public corporations: While the government reviews PC business plans and receives quarterly reports on their financial performance, public reports do not discuss the potential risks to the government’s capital or other potential implications for public finances. Several PCs are loss making, while others receive subsidies for loss-making activities (see principle 3.3.2).

• Government on-lending: The annual debt management report discloses new loans provided over the course of the year, which are on-lent to municipalities and PCs for investment purposes, but information on the stock of outstanding loans and risks surrounding these assets are not discussed.

• Deposit Insurance Fund: The Rulebook on the Investment Policy of the Deposit Insurance Fund governs the management of DIF assets. It states investments must be in either Macedonian Treasury bills or securities issued by the NBRM and requires investments to have a maturity of less than twelve months to ensure liquidity. The Fund publishes an annual report which includes an assessment of market risks associated with the portfolio.

3.2.3. Guarantees (Good)

89. Information on the stock of government guarantees is published quarterly and there are limits on total exposures. The annual debt management report discloses the stock of guaranteed debt by beneficiary. These are solely provided to public enterprises, specifically, PESR, ELEM, MEPSO, MZT, and MDBP. The Law on Public Debt allows the MoF to issue sovereign guarantees on behalf of the government and requires the beneficiary to seek the opinion of the MoF for each loan issued under a guarantee.38 The Fiscal Strategy 2018–20 states that the stock of guaranteed debt must not exceed 13 percent of GDP and provides a limit on the maximum issuance of guarantees in 2018 at 200 million euros. However, these limits are not legally binding.

90. The stock of guaranteed public debt is relatively large and projected to rise further, although their execution has been modest. At end-2016, the stock of government guarantees

38 The Rulebook on the Form and the Contents of the Request for Borrowing by Public Enterprises and Companies Fully or Predominantly Owned by the State, the Municipalities, the Municipalities in the City of Skopje and the City of Skopje sets out the information the borrower must submit to the MoF.

55

was 8.8 percent of GDP, significantly higher than other countries in southeast Europe and higher than the average across Europe (Figure 3.5). Furthermore, the government expects the stock of government guarantees to rise to 11 percent of GDP by end-2019. The government does not publish information on past guarantee calls, but these have been very low in the past, averaging, less than 0.03 percent of GDP a year over from 2013 to 2017.

Figure 3.5. Government Guarantees in Europe, 2016 (Percent of GDP)

Source: Annual Report on Public Debt Management, Eurostat, Fiscal Transparency Evaluations, IMF staff estimates. Note: SEE refers to southeast Europe. Data for Albania is 2014. The figure for Macedonia includes guarantees on borrowing of other public sector enterprises, which to the extent is the case in other European countries, may be not reflected in the data reported to Eurostat.

3.2.4. Public-Private Partnerships (Not Met)

91. The government does not fully disclose its total rights, obligations, and other exposures under public-private partnership (PPP) contracts. PPPs are established under the Law on Concessions and Public-Private Partnerships. The Ministry of Economy (MoE) maintains a public register of 27 concluded PPPs which includes the total value of the contract and concession fee for some of these PPPs. For example, the register does not include one large concession contracts for Skopje and Ohrid airports (total investment is around 2.5 percent of GDP). Further, the register does not include information related to possible direct fiscal obligations or of potential exposures arising from the contract. As such, there is no central understanding of the budgetary exposure to PPPs.

92. There is limited central oversight of PPPs. The Law on Concessions and PPPs requires public entities to undertake a feasibility study prior to contracting PPPs. The MoE, along with MoF (where there are fiscal implications), are required to submit an opinion on the project, but these are generally based on limited project information. The MoE’s opinion primarily focuses on whether the PPP is aligned with the eligibility criteria set out in the law. Robust risk and budget

0

5

10

15

20

25

30

Slov

akia

Czec

h Re

pBu

lgar

iaLi

thua

nia

Latv

iaEs

toni

aIre

land

Rom

ania

Italy

Croa

tiaN

ethe

rland

sAl

bani

aFr

ance

Port

ugal

Gre

ece

Serb

iaPo

land

Spai

nM

onte

negr

oHu

ngar

yU

KM

aced

onia

Cypr

usSl

oven

iaDe

nmar

kSw

eden

Belg

ium

Luxe

mbo

urg

Mal

taG

erm

any

Aust

riaFi

nlan

d

Guarantees (Non-SEE) Guarantees (SEE)Average (Non-SEE) Average (SEE)

56

affordability assessments are typically not conducted, nor is there any central review of the robustness of the feasibility study. Efforts are underway to strengthen the PPP framework, including aligning the PPP law with the 2014 European Union Directive on the Award of Concession Contracts, which would expand the MoE’s oversight powers.

93. While total investment associated with PPPs is high in Macedonia at 6.8 percent of GDP, the total fiscal exposure is lower.39 Many of the projects are concession arrangements and do not appear to have direct fiscal obligations attached to them. This includes the airports concessions, which constitutes close to 40 percent total PPP investment. Other significant projects are the TE-TO Skopje project for electricity generation (30 percent of total investment) and the development of the Technological Development Zone in Tetovo (24 percent of total investment). The remainder consists of a number of smaller projects, primarily entered into by municipalities relating to water and other utilities.

3.2.5. Financial Sector (Advanced)

94. Explicit obligations to the financial sector are disclosed and the NBRM conducts regular assessments of financial stability. The government has direct exposure to the financial system through the DIF, which insures deposits of individual accounts up to a cap of 30,000 euros. The Fund discloses information on the total amount of deposits, the share of insured deposits, and by extension the amount of insured deposits in its annual report. It also provides information on the performance of the Fund. The NBRM publishes an annual Financial Stability Report in July of each year that includes a detailed analysis of financial sector risks. The report also includes aggregated stress test results based on two adverse macroeconomic scenarios. In addition, the NBRM publishes a quarterly report on risks in the banking system, which includes sensitivity analysis of core risk indicators to various discrete shocks.

95. Macedonia’s financial sector is relatively small and well capitalized. Financial sector liabilities were around 74 percent of GDP at end-2016, at the lower end of comparator countries (Figure 3.6). The capital adequacy ratio of the banking sector is more than twice the regulatory minimum, banks have healthy liquidity buffers, and a relatively low share of non-performing loans which are fully provisioned for (Table 3.3). Although, banks have a relatively high maturity mismatch, with close to half the banks’ liabilities that fall due within a month not covered by assets with a similar maturity. Total bank deposits amount to 38 percent of GDP, while insured deposits are close to 29 percent of GDP (Table 3.4). The DIF is required to maintain resources of at least 4 percent of total deposits to meet potential claims. Currently, DIF investments are around 6 percent of total deposits, or 8 percent of insured deposits. Past payouts from the DIF have been relatively low, totaling around 0.2 percent of GDP since 1997.

39 This is based on the data available in the PPP registry for the total estimated value of the investments across all PPPs. Data from the World Bank is used for the two airport concessions and the TE-TO Skopje project for electricity generation, which are not accounted for in the PPP registry.

57

Figure 3.6. Financial Sector Liabilities – Select Countries (Percent of GDP)

Source: Eurostat, NBRM

Table 3.3. Indicators of Banking Sector Stability, 2017

Source: IMF Financial Soundness Indicators.

Table 3.4. Explicit Financial Sector Exposure, 2016 (Percent of GDP)

Source: Source DIF Annual Report and NBRM.

0

200

400

600

800

1000

1200

Mac

edon

ia

Rom

ania

Lith

uani

a

Pola

nd

Slov

enia

Bulg

aria

Croa

tia

Esto

nia

Czec

h Re

p

Latv

ia

Hung

ary

Spai

n

Finl

and

Aust

ria

Port

ugal

Ger

man

y

Belg

ium

Swed

en

Fran

ce

Denm

ark

UK

Asset Quality Profitability FX Exposure

Tier 1 CapitalCapital to

AssetsNPLs Liquid Asset

Loan to Deposits

Return on Assets

FX Liabilities to Total Liabilities

Albania 15.1 10.2 13.2 13.0 191.5 1.5 51.9Austria 15.0 7.5 2.4 23.7 86.9 0.8 5.9Bosnia 14.8 14.0 10.0 28.4 105.1 1.5 55.2Bulgaria 20.9 11.4 10.4 33.1 138.8 1.2 39.4Greece 17.0 12.0 45.6 17.7 68.1 -0.2 4.3Kosovo 16.2 12.6 3.1 28.9 124.4 2.6 4.6Macedonia 14.2 10.8 6.1 23.2 114.0 1.4 45.7Portugal 14.5 7.7 13.3 13.9 87.0 0.3 3.3Slovakia 16.8 10.8 3.7 29.7 101.6 1.1 3.5Spain 13.4 7.6 4.5 14.4 84.4 0.5Sweden 23.4 6.1 1.1 17.5 44.0 0.9

LiquidityCapital

AmountTotal bank liabilities 74.3 Of which deposits 38.0 Of which insured 28.6Total DIF assets 2.3Net exposure 26.4

58

96. Political uncertainty in 2016 saw increased pressure on the exchange rate and relatively large deposit withdrawals, testing the resilience of the financial sector. This led the NBRM to intervene, in April and May 2016, to stabilize the exchange rate and supply banks with foreign currency liquidity of close to 130 million euros, which is over 5 percent of the average amount of foreign reserves during 2016. The relatively large deposit withdrawals posed liquidity challenges for the banking sector, with banks’ liquid assets decreasing by more than 10 percent in the second quarter of 2016. Notwithstanding this episode, the banks remained stable throughout the period and no new drawdowns from the Fund were required.

3.2.6. Natural Resources (Not Met)

97. The government does not publish any estimates of the value or volume of major natural resources, though these are relatively small (Figure 3.7). However, the Geology Department, an independent directorate within the MoE, estimates the volume and value of natural resources on a site-by-site basis every five years as part of a comprehensive study if it is commissioned to do so on the basis of an extraction request. However, these studies are not published. There are 355 concessions for mineral extraction, of which 14 are for metal, 56 for quarrying, 45 for water, and 240 for other non-metal (e.g., sand, gravel, etc.). But, total natural resource revenues derived from these are small, with general government natural resource revenues amounting to less than 0.2 percent of GDP in 2017.

Figure 3.7. Natural Resource Rents, 2016 (Percent of GDP)

Source: World Bank World Development Indicators.

3.2.7. Environmental risks (Not Met)

98. The government does not disclose the main fiscal risks from natural disasters. While the Directorate for Protection and Rescue has carried out a national assessment of risks from

0

5

10

15

20

25

30

Gre

ece

Czec

h Re

publ

icAu

stria

Slov

enia

Port

ugal

Hung

ary

Slov

akia

Turk

ey UK

Pola

ndG

eorg

iaRo

man

iaBo

snia

Bulg

aria

Mac

edon

iaAl

bani

aM

exic

oAu

stra

liaN

orw

ayRu

ssia

Kaza

khst

an Iran

Saud

i Ara

bia

59

natural disasters, it has not been updated since 2007. The Crisis Management Centre (CMC) was set up in 2009 as the National Platform for Disaster Risk Reduction. Currently, it only reports on natural disasters ex post through a monthly bulletin. The CMC does internal risk assessments which are not yet published. It also maintains an internal database on the cost of past natural disasters. In late March 2018, it held a workshop on “Disaster Risk Assessment and Mapping” which is being used to take forward a project to estimate losses from disasters and conduct regional risk assessments. Furthermore, the Ministry of Environment and Physical Planning has a flood management strategy, although it is yet to complete risk-based assessments and risk mapping. The government maintains contingency reserves explicitly for dealing with natural disasters (see 3.2.1).

99. Natural disasters pose moderate risks to the public finances. Macedonia is exposed to various types of natural hazards, including earthquakes, wild fires, floods, droughts, extreme temperatures, landslides. Earthquakes pose the largest risk in terms of impact. For example, the earthquake in 1963 destroyed much of Skopje and resulted in thousands of casualties. It is estimated to have caused economic damage of 15 percent of GDP. Wildfires are a most frequent risk, and floods are on the rise in terms of frequency and intensity. For example, there were significant floods in 2016, resulting in fiscal costs of 0.1 percent of GDP from the amounts disbursed through the contingency reserves. Nonetheless, natural disaster events do not consistently pose a large risk in financial terms, relative to comparator countries (Figure 3.8).

Figure 3.8. Average Annual Damages from Natural Disasters (Percent of GDP)

Source: Global Assessment Report on Disaster Risk Reduction, 2015.

3.3. Fiscal Coordination 3.3.1. Subnational Governments (Not Met)

100. The financial performance of municipalities is published quarterly, but information on the stock of debt, required to assess the financial position of individual municipalities,

0.0

0.5

1.0

1.5

2.0

2.5

Port

ugal UK

Spai

n

Cypr

us

Fran

ce

Pola

nd

Bulg

aria

Czec

h Re

publ

ic

Turk

ey

Aust

ria

Mac

edon

ia

Aust

ralia

Bosn

ia

Rom

ania

Slov

enia

Italy

Alba

nia

Slov

akia

Geo

rgia

Hun

gary

Japa

n

Gre

ece

Earthquake Flood Other

60

is not disclosed. Since the first quarter of 2017, the MoF has been publishing quarterly budget execution reports on the revenues and expenditures of individual municipalities. While the MoF reports on the total stock of commercial borrowings for all municipalities in its quarterly debt bulletins, it does not publish disaggregated information by individual municipality or information on local government liabilities to the central government. Nor does it regularly publish information on broader liabilities, such as stock of arrears. That said, the MoF receives and monitors information on the stock of debt and arrears by individual municipalities on a regular basis. Publishing complete information on total liabilities by municipality along with regular assessment of their financial health would be a relatively straightforward step and would see Macedonia meet advanced level practice under the Code.

101. Fiscal risks are mitigated by strong controls on local government borrowing. The Law on Financing of the Units of Local Self-Government establishes both short-term and long-term borrowing limits, and all borrowing requires approval from the Ministry of Finance. Both limits state that the stock of debt associated with either short-term or long-term borrowing may not exceed 30 percent of total revenues in the current operational budget of the municipality in the previous fiscal year. Further, commercial banks are required to report all loans to municipalities to the MoF.

102. Local governments’ spending and debts are low. Local government expenditure is less than 5 percent of GDP, and they are highly reliant on central government transfers, with own source revenues making up only 34 percent of their funding (Figure 3.9). Furthermore, local government debt remains relatively low at around 0.5 percent of GDP. The vast majority of this is central government on-lending from international financial institutions. Most municipalities are operating with a budget surplus and have low debt to revenue ratios, although a small number ran large deficits to revenue in 2017 (Figure 3.10). The unverified data on arrears highlighted earlier suggests that payment arrears for municipalities could be around 0.8 percent of GDP.

Figure 3.9. Size and Self-Reliance of Sub-National Governments, 2016

Source: IMF GFS. Note: Self-reliance is LG own source revenues as a share of total LG expenditure.

0

10

20

30

40

50

60

70

80

0

5

10

15

20

25

30

Cypr

us

Turk

ey

Alba

nia

Gre

ece

Geo

rgia

Mac

edon

ia

Bosn

ia

Port

ugal

Slov

akia

Hung

ary

Bulg

aria

Slov

enia

Rom

ania UK

Czec

h Re

publ

ic

Fran

ce

Pola

nd

Italy

Aust

ria

Spai

n

Ger

man

y

Swed

en

Belg

ium

Perc

ent o

f Exp

endi

ture

Perc

ent o

f GDP

Subnational Government Expenditure (LHS)Self Reliance (RHS)

61

Figure 3.10. Sub-National Government Financial Position, 2017 (Percent of Revenue)

Budget Balance

Debt

Source: Ministry of Finance.

3.3.2. Public Corporations (Basic)

103. Transfers between the government and PCs are regularly disclosed, but these are not based on a published ownership strategy and there is no report on the sector’s performance. Transfers between the government and PCs, through subsidies and dividend income, are disclosed in the budget. The debt management report discloses the guaranteed debts of PCs by beneficiary, which comprise the bulk of their debt liabilities (Figure 3.10). PCs are required to publish annual financial statements, but many are not up-to-date on company websites or contain qualifications by the auditor. However, the Law on Public Enterprises was amended in April 2018 to require PCs to publish annual accounts on their website within 15 days from the data it has been submitted to the Central Registry. PCs are also required to submit quarterly financial reports to the Government. PCs generally do not have clear performance criteria that they report against, and there is no published document that outlines the purpose and objectives of state ownership and the criteria under which PCs operate.

104. Some PCs in Macedonia engage in a range of non-commercial activities for public policy purposes, for which they may receive subsidies. The Budget reports both subsidies for public corporations and capital subsidies for enterprises and non-government organizations. For example, central government transfers to Macedonian Railways Transport, Macedonian Railways Infrastructure, and the Public Transport Company Skopje (JCP) constitute on average over 0.5 percent of total annual expenditure from 2010 to 2016. These transfers compensate the PCs for losses or for carrying out specific tasks (for example, JCP is compensated for providing free transport for students and pensioners).

105. PCs in Macedonia create some fiscal risks (Figures 3.11 and 3.12). There are around 139 enterprises that are currently treated as being outside of general government by the

-35

-30

-25

-20

-15

-10

-5

0

5

10

15

Ordered Municipalities0

10

20

30

40

50

60

70

Ordered Municipalities

62

authorities, with combined non-equity liabilities of about 15.4 percent of GDP at end-2016.40 Of these, around 1.2 percent of GDP are loans owed to the government. Total liabilities of PCs are relatively low compared to other European countries. Nearly all PC borrowing is backed by a sovereign guarantee. Of total borrowings of 9 percent of GDP, only 0.2 percent of GDP is not guaranteed. The government includes these guarantees in its calculation of ‘public debt’ and so these exposures are already taken into account as part of the government’s debt limit. PCs are also required to obtain approval from the MoF for their guaranteed borrowing. Prior to granting approval for guaranteed borrowing, the MoF reviews the balance sheet, projected revenues and an analysis of the economic, financial, and social benefit of the project the PC is seeking to finance. Non-guaranteed borrowing requires approval from the shareholding ministry.

Figure 3.11. Liabilities of Public Corporations in Select Countries (Percent of GDP)

Figure 3.12. Stock of Guaranteed Public Debt by PC (percent of GDP)

Source: Eurostat, MoF, IMF staff estimates.

3.4 Recommendations

106. Macedonia discloses information on some risks to public finances, but there is scope to enhance the analysis, reporting and management of them. As Table 3.5 shows, the disclosure and analysis of general macroeconomic, guarantees, and financial sector risks meets good or advanced practice. Macedonia also pays considerable attention to risks around conventional debt liabilities and there are strong controls in place for local government borrowing. But, information on fiscal risks is scattered across different documents, and no report

40 This figure includes the PESR and other enterprises that in Chapter I have been classified inside the general government sector for the purposes of providing a public sector overview in line in line with international standards. However, these enterprises are currently classified as PCs in the authorities accounts. Classifying these entities inside the general government sector, has the effect of increasing general government liabilities and reducing PC liabilities. Excluding these entities, PC liabilities (other than equity) would be 8.6 percent of GDP on a non-consolidated basis.

0 10 20 30 40

LithuaniaRomania

CroatiaBulgariaEstonia

HungaryMacedonia

GeorgiaAlbania

LatviaTajikistan

Poland

0

2

4

6

8

10

2009 2010 2011 2012 2013 2014 2015 2016

PESR ELEMMBDP MEPSOMZT

63

provides a comprehensive picture of the government’s aggregate fiscal risk exposure. There is also scope to improve oversight and control of fiscal risks associated with PPPs and PCs.

107. Based on the above assessment, the evaluation highlights the following priorities:

• Recommendation 3.1: Improve fiscal risk analysis and disclosure.

a. Publish a comprehensive statement on fiscal risks comprising:

– Discussion and quantification of macroeconomic and public debt risks building on the existing analysis published in the FSR and ERP;

– All explicit contingent liabilities, including guarantees, obligations under PPPs and the deposit insurance scheme, their magnitude, and where possible, likelihood of realization, and strategies for their management; and

– Analysis of other specific fiscal risks related to public corporations, major legal cases, and natural disasters.

b. Strengthen long-run fiscal sustainability analysis and disclosure by including projections for general government debt and analysis of the sensitivity of the projections to alternative macroeconomic and demographic scenarios.

• Recommendation 3.2: Strengthen oversight and monitoring of PPPs and PCs.

a. Improve disclosure and oversight of PPPs and concessions by:

– Expanding the PPP Registry to record information on direct financial obligations and contingent exposures in contracts; and

– Amending the Law on PPPs and Concessions and by-laws to require that the Ministry of Economy and Ministry of Finance be responsible for assessing the budget affordability and fiscal risks associated with individual projects at all stages of the decision-making cycle, to help safeguard fiscal sustainability.

b. Strengthen the monitoring and assessment of PC performance and risks by compiling a central database of key financial information from PC quarterly and annual reports and defining core risk metrics to monitor their performance.

64

Table 3.5. Macedonia: Summary Evaluation: Fiscal Risks Principle Rating Importance Rec

3.1.1 Macroeconomic Risks

Good: The ERP includes macro-fiscal sensitivity and scenario analysis, but no probabilistic fan charts.

High: Volatility of growth in nominal GDP and revenue is 3.8 and 6.7 percentage points respectively.

3.1.2 Specific Fiscal Risks

Not Met: No report summarizes specific fiscal risks, though relevant information is largely disclosed across various reports.

High: Specific fiscal risks are estimated to be in the order of 46 percent of GDP.

3.1

3.1.3 Long-term Fiscal

Sustainability

Basic: Long-run fiscal projections are published annually in the ERP, but not for multiple macroeconomic or demographic scenarios.

Medium: A rapidly ageing population is estimated to increase the pension deficit by 10 percent of GDP.

3.1

3.2.1 Budgetary Contingencies

Good: The budget includes two contingency reserves for unexpected events with relatively clear criteria for its use.

Low: Use of the reserves amounts to less than 0.1 percent of expenditure.

3.2.2 Asset and Liability

Management

Basic: Borrowing is authorized by law and risks around debt are disclosed and analyzed. Information on some financial assets is disclosed and their risks analyzed.

Medium: General government liabilities and financial assets are 51.5 and 36.5 percent of GDP respectively.

3.2.3 Guarantees Good: The stock of guarantees by beneficiary are regularly disclosed and the FSR sets non-binding limits on their issuance.

Medium: Debt guarantees are 9 percent of GDP, but guarantee calls have been low at less than 0.03 percent of GDP a year on average.

3.2.4 Public-Private Partnerships

Not Met: Government maintains a central registry of PPP contracts, but this does not include the total rights, obligations and exposures under the contracts.

Medium: 19 identified PPPs and concessions, with a total capital stock of 6.8 percent of GDP, but many do not have direct fiscal obligations.

3.2

3.2.5 Financial Sector Exposure

Advanced: Explicit obligations associated with deposit insurance are disclosed, and the NBRM publishes annual financial stability assessments which includes stress tests.

Low: Net exposure for insured deposits is 26.4 percent of GDP, but the banking system is well capitalized and has remained resilient to recent stresses.

-

3.2.6 Natural Resources

Not Met: There is no annual estimate published of the value and volume of natural resources.

Low: Natural resource revenues are less than 0.2 percent of GDP.

3.2.7 Environmental Risks

Not Met: The government does not discuss or quantify fiscal risks from natural or man-made disasters, but has strategies in place to manage these.

Medium: The average annual cost is moderate at 0.3 percent of GDP, but there have been large and costly disaster events.

3.1

3.3.1 Sub-national Governments

Not Met: Revenue and expenditures for individual municipalities is published quarterly, but no report discloses their debt or other liabilities.

Medium: Municipality spending and debt is low at 5 and 0.5 percent of GDP. But, arrears could add a further 0.8 percent of GDP to liabilities.

3.3.2 Public Corporations

Basic: Transfers between PCs and the government and their guaranteed debt is published regularly, but there is no consolidated report on PC performance and no ownership policy in place.

Medium: PC debt is relatively modest at 9 percent of GDP, but almost all of it is backed by government guarantee, creating explicit exposure for government.

3.2

Fiscal Affairs Department

International Monetary Fund700 19th Street NWWashington, DC 20431USAhttp://www.imf.org/capacitydevelopment